一文速学数模-K-means聚类算法实战：信用卡用户画像聚类分析

2023-04-22

数据用户画像

目录前言一、用户画像概述1.用户画像2.为何用聚类算法作用户画像二、数据质量校验1.数据背景2.数据说明三、数据预处理1.数据空缺值检验 2.数据归一化四、K-means聚类step1:选取K值手肘法step2:计算初始化K点step3:迭代计算重新划分五.画像分析前言该项目算是非常经典的

前言

一、用户画像概述

1.用户画像

2.为何用聚类算法作用户画像

二、数据质量校验

1.数据背景

2.数据说明

三、数据预处理

1.数据空缺值检验

2.数据归一化

四、K-means聚类

step1:选取K值

手肘法

step2:计算初始化K点

step3:迭代计算重新划分

五.画像分析

前言

该项目算是非常经典的金融业务用户画像的基础分析了，主要根据用户信用卡使用行为数据进行分析，根据收集到的不同字段信息，对每个用户划分类别。这里需要说明一下的是，聚类模型只是将具有相似行为的大部分用户聚集到一个类别里面，这点并不会考虑到每个字段的含义，也就是分成的类别并不是用户价值等级，此类别仅仅是这个类别大体相同的信用卡用户行为对象，并不能给每个用户价值打上标签，那是评价模型做的事，这里要注意不要搞混淆了，用户价值评价是评价模型决定的，而画像是聚类模型分析的。

博主专注建模四年，参与过大大小小数十来次数学建模，理解各类模型原理以及每种模型的建模流程和各类题目分析方法。此专栏的目的就是为了让零基础快速使用各类数学模型以及代码，每一篇文章都包含实战项目以及可运行代码。博主紧跟各类数模比赛，每场数模竞赛博主都会将最新的思路和代码写进此专栏以及详细思路和完全代码。希望有需求的小伙伴不要错过笔者精心打造的专栏。

以下是整篇文章内容。

一、用户画像概述

数据分析主要是根据用户行为以及不同的业务情况去分析，而用户画像的需求是我们首先需要了解的建模场景。什么是用户画像？为什么要建立用户画像，怎么去建立？这需要我们先了解这一点，才好方便建模。

1.用户画像

用户画像是指根据用户的属性、用户偏好、生活习惯、用户行为等信息而抽象出来的标签化用户模型。通俗说就是给用户打标签，而标签是通过对用户信息分析而来的高度精炼的特征标识。通过打标签可以利用一些高度概括、容易理解的特征来描述用户，可以让人更容易理解用户，并且可以方便计算机处理。简单来说就是：利用已经获得的数据，勾勒出用户需求、用户偏好的一种运营工具和数据分析方法。

根据用户画像的作用可以看出，用户画像的使用场景较多，用户画像可以用来挖掘用户兴趣、偏好、人口统计学特征，主要目的是提升营销精准度、推荐匹配度，终极目的是提升产品服务，起到提升企业利润。用户画像适合于各个产品周期：从新用户的引流到潜在用户的挖掘、从老用户的培养到流失用户的回流等。

2.为何用聚类算法作用户画像

聚类算法是一种常用的无监督学习算法，可以将具有相似特征的数据点归为一类。在用户画像中，聚类算法可以将相似的用户归为一组，从而为营销策略、产品设计和推荐系统等提供有价值的信息。

以下是使用聚类算法作用户画像的一些好处：

识别用户群体：聚类算法可以将用户数据聚类成不同的群体，这有助于识别不同的用户群体，为营销策略提供有价值的信息。
了解用户偏好：聚类算法可以将具有相似特征的用户聚类在一起，从而了解用户的偏好和需求。这可以帮助企业根据用户的喜好和需求开发相应的产品或服务，提高用户满意度。
个性化推荐：聚类算法可以将用户划分为不同的群体，为个性化推荐提供有价值的信息。基于用户画像，企业可以向用户提供更加符合其喜好和需求的产品或服务，从而提高用户满意度和忠诚度。
精细化运营：基于用户画像，企业可以根据用户的需求和行为开展精细化运营，提高用户留存率和复购率。

二、数据质量校验

在了解了上述用户画像的作用之后，我们需要对我们拿到的数据进行分析，这里我采用的是kaggle信用卡用户行为数据。

1.数据背景

信用卡营销就是指通过激发和挖掘人们对信用卡商品的需求，设计和开发出满足持卡人需求的信用卡商品，并且通过各种有效的沟通手段使持卡人接受并使用这种商品，从中获得自身最大的满足，以实现经营者的目标。对于不同的客户群体，需要采用不同的营销手段，因此需要对客户进行分组聚类。

2.数据说明

该文件位于具有18个行为变量的客户级别。

CUST_ID：信用卡持有人的身份证明（分类）
BALANCE：他们帐户中剩余的余额可用于购买
BALANCE_FREQUENCY：更新余额的频率，得分在0到1之间（1 =频繁更新，0 =不频繁更新）
PURCHASES：通过帐户购买的金额
ONEOFF_PURCHASES：一次性购买的最大金额
INSTALLMENTS_PURCHASES：分期购买的金额
CASH_ADVANCE：用户预先提供的现金
PURCHASES_FREQUENCY：进行购买的频率，得分在0到1之间（1 =频繁购买，0 =不频繁购买）
ONEOFFPURCHASESFREQUENCY：一次购买的频率（1 =频繁购买，0 =不频繁购买）
PURCHASESINSTALLMENTSFREQUENCY：分期购买的频率（1 =频繁执行，0 =不频繁执行）
CASHADVANCEFREQUENCY：多长时间支付一次预付款
CASHADVANCETRX：使用“高级现金”进行的交易数量
PURCHASES_TRX：进行的购买交易数
CREDIT_LIMIT：用户的信用卡限额
PAYMENTS：用户完成的付款金额
MINIMUM_PAYMENTS：用户的最低付款金额
PRCFULLPAYMENT：用户支付的全额付款的百分比
TENURE：用户的信用卡服务的使用权

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">

为了更加直观的去作用户画像，这里我仅选择偏业务的正向数据进行建模，当然聚类并不是挖掘客户的消费潜力价值，我这里选择正向特征仅是为了更直观的进行画像描述，除去偏业务方向的知识，仅对画像建立过程优化。

根据特征字段，我们需要选定建模需要的字段信息，若为正向指标则越大越好：

1.BALANCE：帐户中剩余的余额可用于购买，那么该数值越大，用户购买能力越大，正向指标 1

2.BALANCE_FREQUENCY：更新余额的频率正向指标 1

3.PURCHASES：通过帐户购买的金额正向指标 1

4.ONEOFF_PURCHASES：一次性购买的最大金额正向指标 1

5.INSTALLMENTS_PURCHASES：分期购买的金额正向指标 1

6.CASH_ADVANCE：用户预先提供的现金正向指标 1

7.PURCHASES_FREQUENCY：进行购买的频率，得分在0到1之间（1 =频繁购买，0 =不频繁购买）正向指标 1

8.ONEOFFPURCHASESFREQUENCY 一次购买的频率（1 =频繁购买，0 =不频繁购买）正向指标 1

9.PURCHASESINSTALLMENTSFREQUENCY：分期购买的频率（1 =频繁执行，0 =不频繁执行）中向指标与上一个指标互斥

10.CASHADVANCEFREQUENCY：多长时间支付一次预付款负向指标0

11.CASHADVANCETRX：使用“高级现金”进行的交易数量正向指标 1

12.PURCHASES_TRX：进行的购买交易数正向指标 1

13.CREDIT_LIMIT：用户的信用卡限额正向指标 1

14.PAYMENTS：用户完成的付款金额正向指标 1

15.MINIMUM_PAYMENTS：用户的最低付款金额中性指标

16.PRCFULLPAYMENT：用户支付的全额付款的百分比正向指标1

数据差异很大，我们尽可能拿出所有可用到的正向指标聚类，那么高纬度的用户群体就是重要价值用户。

三、数据预处理

1.数据空缺值检验

df_data.isnull().sum()

CUST_ID                               0
BALANCE                               0
BALANCE_FREQUENCY                     0
PURCHASES                             0
ONEOFF_PURCHASES                      0
INSTALLMENTS_PURCHASES                0
CASH_ADVANCE                          0
PURCHASES_FREQUENCY                   0
ONEOFF_PURCHASES_FREQUENCY            0
PURCHASES_INSTALLMENTS_FREQUENCY      0
CASH_ADVANCE_FREQUENCY                0
CASH_ADVANCE_TRX                      0
PURCHASES_TRX                         0
CREDIT_LIMIT                          1
PAYMENTS                              0
MINIMUM_PAYMENTS                    313
PRC_FULL_PAYMENT                      0
TENURE                                0
dtype: int64

空缺处理方法有很多种，大家可以去看我的另一篇文章：

一文速学(四)-数据分析Pandas处理缺失值操作各类方法详解

这些数据基本全为定量数据，可以直接使用机器学习算法对空缺值进行填充，这里使用随机森林填补：


# 随机森林填补缺失值
from sklearn.ensemble import RandomForestRegressor   
RF = RandomForestRegressor()
paymentsNotNull = df_data.MINIMUM_PAYMENTS.notnull()
columns = ['BALANCE','BALANCE_FREQUENCY','PURCHASES','ONEOFF_PURCHASES','INSTALLMENTS_PURCHASES','CASH_ADVANCE','PURCHASES_FREQUENCY','ONEOFF_PURCHASES_FREQUENCY','PURCHASES_INSTALLMENTS_FREQUENCY','CASH_ADVANCE_FREQUENCY','CASH_ADVANCE_TRX','PURCHASES_TRX','PAYMENTS']
RF.fit(df_data.loc[paymentsNotNull,columns],df_data.MINIMUM_PAYMENTS[paymentsNotNull])
print(RF.score(df_data.loc[paymentsNotNull,columns],df_data.MINIMUM_PAYMENTS[paymentsNotNull]))

分数还是不错的，可以达到90.8%：


# 填充
for data in [df_data]:
    pred = RF.predict(data.loc[:,columns])
    IsNull = data.MINIMUM_PAYMENTS.isnull()
    data.MINIMUM_PAYMENTS[IsNull]=pred[IsNull]

这样就填充完毕了：

2.数据归一化

这些数值类型数据最好进行归一化，关于归一化的所有处理步骤都有在我的另一篇博客整理：

一文速学-数据预处理归一化详细解释

我这里采用的z-score作为归一化函数：


from sklearn.preprocessing import StandardScaler  # 标准化工具
scaler = StandardScaler()
for n in range(1,14):
    x_train = scaler.fit_transform(df_true_data.iloc[:,n:n+1].values)
    df_true_data.iloc[:,n:n+1]=x_train

得到归一化数据集，那么开始聚类建模。

四、K-means聚类

K-means的基础原理在我专栏里面有详解，需要温故一下的可以看这篇文章：

一文速学数模-聚类模型(一)K-means聚类算法详解+Python代码实例

具体建模框架以及流程：

QQpHccDr2A9avEU1KSim7vz6dSaUmlfoin/angv8A6F67/wDAlv8A4qmeONL03Sb+zt9Ptmg8y3E0gLlupIHJPsam1XVtJ8T6dEzWf2fX3lSPdEMJJnjc3t+vTqM0fER/tHi/7PCGd4oY4toBJyeQB6/eFFO6qR3W97u/YJW5Ht0N7RfDs8Xg42mp6bbzF737Qkc155S7fLAD7kz7jHvUuqaZC3hS0g0+5s9Pkjvcw+TeNIpkwRtEhwQefwxVXSLC50fRt+t2diZuRaJNai4uWOBtVQCOBjOM9/4cVsaTeats1W31VLLTJoLeOSKaGJdse/d8x5IOCP0Nck5S5nJO9n8vu1N4pWtYhuNSj03To7bWZLi4uJIWgn1W2t/lgVui78c8+np6153NFJoniBU0q/jvJYmHkzwLu3MRwAOQTz2zXa6rc3Ok6a6eKb6fU/tKYjtreLZDnkgmTaOeAeOR1wa8/wBMltYNSglvY5JLdGy6xNtY8cYPGOcV04aNoye/5P8AL/IxrPVI9FsLvxJZRjUPE2s/YbNDkQFI/Nm4zgADj+fXp1q1f+M3TQ9L1lY3S3nu3WSFWG5oxuA5I68A/pnvXN3GteFNTu/MuNL1a5uJDtBacsx54A+b36Cuqu9O0P8AsRdINhPK9rGLz+z45f3wDE55z8xGTkA+nqM804xTTnH8Elb7/wAWaxbaaizN1XUNauNNl1jw9rxubBAxmhkjjEkOOo5Xnj15+tYPw1bb4sx/et3H8jS/2n4OhWWEaRqkYf5ZE88jODnBG7sazfDWu22ga/Jfm3kkhKOiICNwBIx19hXTGm/ZTgo/gl+Rk5LnjJst2PjbxFNqFtFJqTFHlVWHlJyCR/s1t+N/FGs6T4ja1sb0wwiJW2iNTyfqKzbPx9qs99bwtb2AWSRVJEHOCcetbPjLxbf6Nr7WltDaPH5atmWLccn3zUSh+9S9mtn+nkUpe4/eZR8YXU994I0G6uZPMmlYs7YAyce1YZ8P2sfgYa5LPILmWfy4oxjaRnH1zwx61J4i8W/8JBpFjavaiKeBi0jKcKT04FTeI9UspPCuh6Xp88cqwqXn2gjEmB2OO7P25rWnGpCMY2tq/u1Ik4ybe+g3xHZW+j2OgXumo1rcXNoJZHjlfJYqvTJ46np61n+GbLT9W1xbTVLiSNZ1ZY3VuTIemSfx+pxWx40/5Afhb/rwX/0FK53TNJ1DUhNLp8TSNbAO+1gCo55H5fWrpu9G7dt9fmTLSpZIZq2l3Gj6nNY3KkPG2AccOvZh7GqVa+veIbnxDNby3UUKvDGI9yLhm9ST9cnHbNSxeFNRm8OnXFMH2QKzYLndgMVPGPUVqp8sV7TR/qQ43b5NjDooorUg0NC/5GHTP+vuL/0MV6D4hTWjrt0bXwvp15Blds81oHZ/lHU5554/CvPtC/5GHTP+vuL/ANDFehXNlrN58QLgs1+ujRupkK3DRRgeWOhyM89cfpXBiXaom7aJ7nTS+C3maM2f+E+GVCn+wzkAYx+8NcV4M/5BXib/ALBr/wDoLV0el6Zd6Z45vFuLma6gk0+RraaRy+Y9y/Lk+mf1z3rK8MaVPY+Fda1KZ4fs13p8ixsr5O7DDBHrnisYuMYNX/l/M0d3JO3craYrP8LNWVQWY3qAADJJzHWVq3hxdE0y3lvb0LqE43CyWPJRfVmzx+XX6HHS+ENQi0rwNe309uJ4ob9GZCB/0z5Ge46j3FYvjfSZLfUhq0U73VlqB8yOYnOCf4c/Tp7DHat6c37Zx2V389FoZyiuRPyL1joOh6DZQ6rr17FdmRd9vaW53B+/PqPrgdjmltbtPF51y/1OAP8AY7NpLSIOwSLAY44Iz0GTXD113gz/AJBXib/sGv8A+gtVVabjFzbu/wAtehMJJtRS0LGm+INM16wj0rxSG3xcW9+D8y54+Y/1OQcc9M1i6/pUHh/UoBY6lHebgJkdF+6M/Lk8g/59aw6ktzEtzE06l4Q4MiqcErnkD8K1jR5JXi9O3Qhz5lZ79z0bTZ/F7x/bdZ1c6Zp64ZpJoow7D0VccH6+vQ1o3fjFrjwxeapYh0jtrtIUyQGkQbd3UHGcntxXM3WveE77y/tWnarN5ahU33JO0AAYHzcdBXV2mn6ENBXTPsM0B1BWuUsZZcSybMdyeDwDjI7+hx59SMU1Kcba9klbr1OqDbuosqT6nqmvWj3/AIX1ptyDM1jOiBk47Ejn8Tj37VxfhO4km8cWNxK2+WWdmdsYyWByf1rSN/4RsZpIm0XVreXBjkUzFGweoI3d/Ssaw1Sw0zxVHqVtbSiyikLJCzZYDGOv1rop07QlGMd1povzW5lKV5JtmtrXjPxBa67qFvDqLJFFcyIi+UhwoYgDpW7418SavpX9l/Yrwxedbb5PkU7jxzyKxLn4i6k91M0FrZCJnJQSQ5YLnjJzycV0Pi/xRfaN/Zv2eK1fz7fe3mxbsHjpzWbptTgnBdfnp6FqV4yfMzO1DUrvVfhabq9mM0xugCxUDgNx0rn/AAr4ftdaXUZr2eSG3s4PMJjwDnk8k9sA1a1Xxq+seGm064tES4aYOZIvlTaPb1o0jVLLTvAmrwJPH/aN0wTyyDnyzgdeh4L9DxmtYxqQptJWbfQhuMppt3siG10uxl+Ht7qbW4+2xXYiSXe33fk7Zx/Ee1Yaz/a7+F9RuJnjLqsshYs4TPOM+1dRY/8AJKNS/wCv8fyjrlbGxn1K9jtLVN88mQi5xk4zj9K2pP43J7N/oZzXw27Gz4r8PLol3FNZs0um3Sh7eXORz/Dn9R7H61qzXdzZ/DDTJLW4lgc3rqWicqSPn4yKx9Q1jUrfRR4avreMfZpc5kG507gA5wOp/A4rRv8A/klml/8AX8//ALPWbUuWCnrr9+5SteTj2NKfQtSszGt342W3kdA4SW5dTg9+W+tZXiGHWtAjs5P+EiubqO7VmR4p3xgY5685zV7xFP4b8RXlvdNr7Wxit1hKfYnfoSc549f0rP8AFup6beadotnp12bkWULRO5iZM8KAcH1wazpc7lHmXrp+tip8qTt+Zc8d/ZP+E5X7d5/2XyU8zyMb8YPTPHWtjRYg+jRjShrTaeXYoHhs2G7PP3zmuf8AiR/yNr/9cE/rW5oLI/g2zsp7GZmWRpQZtKkuYyCTgjGB0PXNRJf7PB+hS/iyNL7Le4z5Wq49fs1h/jVDWogmjyDVf7aXTw6lwkNmo3Z4+4c1qfZof+Ee8v7FDt+158v+w5MZ2dfK3Zz/ALecdqx9fZI/Bt5Zw2UqM0iyExaVJbRgAjJOcjPHXNY03eS9expJWT/zPPbv7L9rk+xed9mz+787G/HvjioKKK9haI4T1bwF/wAixH/11f8AnRR4C/5FiP8A66v/ADor5vEfxperPVpfAjg/Fv8AyNeo/wDXX+grFra8W/8AI16j/wBdf6CsWvoKH8KPojzKnxv1NvSPFusaJB5FpcjyMkiKRQygn07iq+r6/qeuSK1/cmRVJKIAFVfoB/8ArrMopqlBS5rai55WtfQ29K8Waxo1k1pZ3IEJJKq6BthPdc9P5VU0nW9Q0O4aawuDEXADrgFXA9Qf/wBfJrPop+zhrpvuHPLTXY3dV8XaxrFq1rczoLd8boo4woYg5ye/Yd+1b6a3caR8O9IlsLpY7lbpwy5BJXL8EenSuDorOVCDSilZJ3KVSSbZq6x4j1PXdgvrjdGhykSKFVTjGcDqfr6mrNn4y1+wtIrW2v8AZDEu1F8mM4H1K5rBoq/ZU+Xl5VYnnle9zV1XxJq2tQJDqF350aNuUeWi4OMfwgVDca1qF3pkGmz3G+0tzmKPYo2nkdcZPU9TVCimqcErJbA5Se7NfXtUg1T+zPISRfsthFbPvAGWXOSME8c1of8ACbXkNtbWllZWlvZRRbJLYxh0mJHJbPPP1+pNcxRUujBpJrYfPK90dWnirSGVTceE7B3x8xifywfoMHFZWt6pY6m8JsdJi05YwdwR928nHU4HpWTRRGjCLuvzYOpJqzOgstetrbwhcaNLbySSS3az5zhSo2ZBOc/wn86j8SeIP7dmtVitxbWlrCI4oAQQpwN38gPoBWHRQqUFLm6g5yasdRpXiy3stHttPvdJW8W1uPtELicx7W6jIAOeSfb2qKHxSn/CUXOt3emw3DScxxsT+7IA2kH1GBzj6YrnKKXsKd27bj9pLTyOoi8d6sZG+3JbX0LOH8qeIEJzn5SOR7dcYpJNf8OzNvl8JR7sY/d3roPyArmKKPYU+it6XX5B7SXU62XXfC0NpMth4bxcOuEaeUuqn1wSemfxrCtdb1OxsJLK0vZYLeR97LGdpJ4HUc9h3rPopxpRStv66/mJzbOg8Pazb6faa3HdtIZL2zeKNgM5cg9fz60zVfE02oaTaaXb20dlZQKN8UJJEj/3iTzjvg555JPGMKij2MObme4c8rWN6/8AFNxqegQ6de20M88LYS8fmQJ6fXgZPcds81m3eq3t/aWlrczeZDaKUgXaBsBxnkDJ6DrVOiqjThHZCc5Pdl/+2tQ/sf8Asn7Rmx3bvKKKcHOeDjI59DVCiiqUUthNt7m/p2s2dt4S1TSZ0nM926vGyKNo2kEZJOeo9KwKKKmMFFtrqNybsaWg31rputWt5eQPPDC+7YpGc9jz1wea0IvFUy+M/wC3pA5UyHMYP/LPGNv5frXO0UpUoybb6qw1NpWR0uleKY9N8TXWpLaFbO73+dah93B568Dr7dCRTfD/AIlh0e5vopbRpdMvFYNbb87eu3k9fQn8e2K5yipdCDvpv+gKpJE9tMkF/DPtby45Q+3OTgHPtzWr4s1q317W2vraOWOMxqmJAAcj6E1h0VbgnJS6oXM7WOhuNWs38CWmkozm7jvDM4K8BcMOv4iueooojBRvbqEpN7nS6P4oh0PRJYrKwRdWkJX7Y3OEP17j06dzmuckkeWRpJHZ5HJZmY5LE9ST602iiNOMW2t2Dk2kmdToOseGtMtLeW70ea41KFywlDnaTnKnG7HHHbtWJrOpyazq9xqEqKjTNnavRQAAB+QFUaKUaUYyc+o3Ntcp10XjRbTT1e1tGbWXQpLfXD+YVH+xnoPbgD3qlpfiNbaw1yO9M89zqMQRZODz83LZPuOlc9RU+whZq24/aSOg0Xxjqejw/Zspd2eRmC4G4Aeint/L2qpPf2GoeIFup7FbWwZ1D29thdq98EDn16flWVRVKlBNySs2LnlazOx/4SvR9IU/8I5owjnI/wCPm7O51+gyf51zQ1S+Gp/2kLqT7bv3+dn5s/4Y4x0xxVOilClCN/PuEpyZ2T+K9H1eBj4g0USXar8txanYXPON3I/mfpXG0UU4U4w+EUpuW5PZypb31vNIGKRyq7BeuAcnFavizWrfXtba+to5Y4zGqYkAByPoTWHRTcE5KXUOZ2sFFFFWSdgnjSwl0uxs9R8PQ3jWkQiSRp8cAAdNpx0Hemz+M7NNKvbLTNAgsWu4/LkkWXd8p4PG0die/euRorD6tT7fizT2swrqrPxELTwNd6ZJeiaWY+XBaiLHkqWJZi2BnOeBk4498crRWk6amkn01JjJx2CiiirJLemX8ul6nb30IUvA4cBhkH1H5Vd1nxPq2uMftlyfJPSCP5Yx+Hf8c1j0VDhFy5mtSuZpWOr0DxjLYzRrqTzTWsFo9vAkSjIJIwTkjPAxXMGeVoVhaVzEpyqFjtB9hUdFKNOMW2luDm2rMuR6rexaVNpiTYs5nEkke0csMc5xnsO9OXWdQXSH0oXBNizbzEyg4Oc8EjI5HY+vqao0VXJHsLmYVv8AhvWbPSrbVYbxJ2F7bGBTCoOMgjJyR61gUUpwU48rHGTi7oKchUOpcFkz8wU4JHsecU2irJOyHiLw3o+ZNE0VpbofdnvGyFPqBk/0rmr7Vr7UdSOoXNw7XOQVcHGzHI2+mPaqVFZQpRi77vzLlNvQ7GHxhYanAsHifS1vCowtzANsvbryP0I+lclcNE9zK0CGOEuTGhOSq54GfpUdFOFKMH7oSm5bhXR+LNftddaw+yxTILaHy280AZPHTBNc5RTcE5KT6CUmk13CiiirJOl0HxVBpOjXGl3elR39vNL5pV5NvOAPQ/3RV+18aaPY3C3Fp4Ut4p0zscXHK9v7lcXRWEsPTk22t/NmiqyWxNd3Ml5eT3UuPMmkaR8DjJOT/OtW41uKbwhZ6KIXEsFwZjISMEHdx/49WJRWjgnbyJUmr+YVteHtS0vTpZzqmmLfRuo2DjKsD79iD+grFopyipKzEnZ3Rs61ro1zXxqVxaARZQGASHlR1G7HfnnHetVPF2mRxrGmjXqxoMKq6vMAo9AMVyNFZujBpLovNlKpJO523/CdWX2L7J/Y915Xmebn+1ZN2cY+9jOPbOKryeLdLliaKTRbx43GGVtXmII9xiuRoqVhqa2/N/5le1n/AEkSTvG9xI8MXlRM5KR7t2wZ4Ge+PWo6KK3Mj1bwF/yLEf8A11f+dFHgL/kWI/8Arq/86K+bxH8aXqz1qXwI4Pxb/wAjXqP/AF1/oKxa2vFv/I16j/11/oKxa+gofwo+iPMqfG/UKKKK1ICiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooA9W8Bf8AIsR/9dX/AJ0UeAv+RYj/AOur/wA6K+axH8aXqz1qXwI4Pxb/AMjXqP8A11/oKxa2vFv/ACNeo/8AXX+grFr6Ch/Cj6I8yp8b9QooorUgKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigD1bwF/yLEf/XV/50UeAv8AkWI/+ur/AM6K+axH8aXqz1qXwI4Pxb/yNeo/9df6CsWuv8SeG9ZvPEV9cW+nyvE8mVYY54FZX/CJa9/0DJv0/wAa9yjWpqnFOS2XU8+pTm5vQxaK2v8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/GtPb0v5l95Hs59mYtFbX/CJa9/0DJv0/xo/wCES17/AKBk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/wAaPb0v5l94ezn2Zi0Vtf8ACJa9/wBAyb9P8aP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/hEte/6Bk36f40f8Ilr3/QMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/xo/4RLXv+gZN+n+NHt6X8y+8PZz7MxaK2v8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/wAaP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/hEte/6Bk36f40f8Ilr3/QMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/xo/wCES17/AKBk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/wAaPb0v5l94ezn2Zi0Vtf8ACJa9/wBAyb9P8aP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/hEte/6Bk36f40f8Ilr3/QMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/xo/4RLXv+gZN+n+NHt6X8y+8PZz7MxaK2v8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/wAaP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/hEte/6Bk36f40f8Ilr3/QMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/xo/wCES17/AKBk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/wAaPb0v5l94ezn2Zi0Vtf8ACJa9/wBAyb9P8aP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/hEte/6Bk36f40f8Ilr3/QMm/T/Gj29L+ZfeHs59mYtFbX/CJa9/0DJv0/xo/4RLXv+gZN+n+NHt6X8y+8PZz7MxaK1YfDWsXHmeTYSv5TmN8EfKw6jrUv/CJa9/0DJv0/xodamt5L7w9nPsYtFbX/AAiWvf8AQMm/T/Gj/hEte/6Bk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8aP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8AGj/hEte/6Bk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8aP8AhEte/wCgZN+n+NHt6X8y+8PZz7MxaK2v+ES17/oGTfp/jR/wiWvf9Ayb9P8AGj29L+ZfeHs59mYtFbX/AAiWvf8AQMm/T/Gj/hEte/6Bk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8aP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8AGj/hEte/6Bk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8aP8AhEte/wCgZN+n+NHt6X8y+8PZz7MxaK2v+ES17/oGTfp/jR/wiWvf9Ayb9P8AGj29L+ZfeHs59mYtFbX/AAiWvf8AQMm/T/Gj/hEte/6Bk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8aP+ES17/oGTfp/jR7el/MvvD2c+zMWitr/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xo9vS/mX3h7OfZmLRW1/wiWvf9Ayb9P8AGj/hEte/6Bk36f40e3pfzL7w9nPszFora/4RLXv+gZN+n+NH/CJa9/0DJv0/xo9vS/mX3h7OfZneeAv+RYj/AOur/wA6Ks+DdOvLHw+sFzbvHKJGJUj1or5+u06smu56lJWgrnZUUUViWFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAGF4Z+9rX/YUm/ktbtYXhn72tf9hSb+S1u1pW+NkQ+EKKKKzLCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAMLwz97Wv8AsKTfyWt2sLwz97Wv+wpN/Ja3a0rfGyIfCFFFFZlhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAGF4Z+9rX/AGFJv5LW7WF4Z+9rX/YUm/ktbtaVvjZEPhCiiisywooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiuf8YtqcOgyXelXLwzW58xwoB3J36jt1/A1UI88lHuKTsrj/DP3ta/7Ck38lrdrx7wjqevalry2cF/IkU0puLohV6cbj04zwPxr2Gt8VSdOdmzKjNTjoFFFFcxsFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAHN3vjnQ9PvJbS4mlWaJtrARE1B/wsXw7/z8Tf8Aflq8z8W/8jXqP/XX+grFr2KeX0pQUm3qv66HDPFTUmrI9m/4WL4d/wCfib/vy1H/AAsXw7/z8Tf9+Wrxmir/ALNpd3+H+RP1ufZHs3/CxfDv/PxN/wB+Wo/4WL4d/wCfib/vy1eM0Uf2bS7v8P8AIPrc+yPZv+Fi+Hf+fib/AL8tR/wsXw7/AM/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/wA/E3/flqP+Fi+Hf+fib/vy1eM0Uf2bS7v8P8g+tz7I9m/4WL4d/wCfib/vy1H/AAsXw7/z8Tf9+Wrxmij+zaXd/h/kH1ufZHs3/CxfDv8Az8Tf9+Wo/wCFi+Hf+fib/vy1eM0Uf2bS7v8AD/IPrc+yPZv+Fi+Hf+fib/vy1H/CxfDv/PxN/wB+Wrxmij+zaXd/h/kH1ufZHs3/AAsXw7/z8Tf9+Wo/4WL4d/5+Jv8Avy1eM0Uf2bS7v8P8g+tz7I9m/wCFi+Hf+fib/vy1H/CxfDv/AD8Tf9+Wrxmij+zaXd/h/kH1ufZHs3/CxfDv/PxN/wB+Wo/4WL4d/wCfib/vy1eM0Uf2bS7v8P8AIPrc+yPZv+Fi+Hf+fib/AL8tR/wsXw7/AM/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/wA/E3/flqP+Fi+Hf+fib/vy1eM0Uf2bS7v8P8g+tz7I9m/4WL4d/wCfib/vy1H/AAsXw7/z8Tf9+Wrxmij+zaXd/h/kH1ufZHs3/CxfDv8Az8Tf9+Wo/wCFi+Hf+fib/vy1eM0Uf2bS7v8AD/IPrc+yPZv+Fi+Hf+fib/vy1H/CxfDv/PxN/wB+Wrxmij+zaXd/h/kH1ufZHs3/AAsXw7/z8Tf9+Wo/4WL4d/5+Jv8Avy1eM0Uf2bS7v8P8g+tz7I9m/wCFi+Hf+fib/vy1H/CxfDv/AD8Tf9+Wrxmij+zaXd/h/kH1ufZHs3/CxfDv/PxN/wB+Wo/4WL4d/wCfib/vy1eM0Uf2bS7v8P8AIPrc+yPZv+Fi+Hf+fib/AL8tR/wsXw7/AM/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/wA/E3/flqP+Fi+Hf+fib/vy1eM0Uf2bS7v8P8g+tz7I9m/4WL4d/wCfib/vy1H/AAsXw7/z8Tf9+Wrxmij+zaXd/h/kH1ufZHs3/CxfDv8Az8Tf9+WpG+Ifht1KtNKVIwQYDgivGqKP7Npd3+H+QfW59kei+Gdc8LeHZb6RLmV2nlPlnyWysY6L9eTn8K6H/hYvh3/n4m/78tXjNFVPAU5y5pN3+X+RMcTKKskj2b/hYvh3/n4m/wC/LUf8LF8O/wDPxN/35avGaKn+zaXd/h/kV9bn2R7N/wALF8O/8/E3/flqP+Fi+Hf+fib/AL8tXjNFH9m0u7/D/IPrc+yPZv8AhYvh3/n4m/78tR/wsXw7/wA/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/z8Tf8AflqP+Fi+Hf8An4m/78tXjNFH9m0u7/D/ACD63Psj2b/hYvh3/n4m/wC/LUf8LF8O/wDPxN/35avGaKP7Npd3+H+QfW59kezf8LF8O/8APxN/35aj/hYvh3/n4m/78tXjNFH9m0u7/D/IPrc+yPZv+Fi+Hf8An4m/78tR/wALF8O/8/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/AM/E3/flqP8AhYvh3/n4m/78tXjNFH9m0u7/AA/yD63Psj2b/hYvh3/n4m/78tR/wsXw7/z8Tf8Aflq8Zoo/s2l3f4f5B9bn2R7N/wALF8O/8/E3/flqP+Fi+Hf+fib/AL8tXjNFH9m0u7/D/IPrc+yPZv8AhYvh3/n4m/78tR/wsXw7/wA/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/z8Tf8AflqP+Fi+Hf8An4m/78tXjNFH9m0u7/D/ACD63Psj2b/hYvh3/n4m/wC/LUf8LF8O/wDPxN/35avGaKP7Npd3+H+QfW59kezf8LF8O/8APxN/35aj/hYvh3/n4m/78tXjNFH9m0u7/D/IPrc+yPZv+Fi+Hf8An4m/78tR/wALF8O/8/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/AM/E3/flqP8AhYvh3/n4m/78tXjNFH9m0u7/AA/yD63Psj2b/hYvh3/n4m/78tR/wsXw7/z8Tf8Aflq8Zoo/s2l3f4f5B9bn2R7N/wALF8O/8/E3/flqP+Fi+Hf+fib/AL8tXjNFH9m0u7/D/IPrc+yPZv8AhYvh3/n4m/78tR/wsXw7/wA/E3/flq8Zoo/s2l3f4f5B9bn2R7N/wsXw7/z8Tf8AflqP+Fi+Hf8An4m/78tXjNFH9m0u7/D/ACD63Psj2b/hYvh3/n4m/wC/LUf8LF8O/wDPxN/35avGaKP7Npd3+H+QfW59kezf8LF8O/8APxN/35aj/hYvh3/n4m/78tXjNFH9m0u7/D/IPrc+yPftN12y1a0F1aM7RFioLLjkUVy/gL/kWI/+ur/zoryKsFCo4rod0JOUU2cH4t/5GvUf+uv9BWLW14t/5GvUf+uv9BWLX0VD+FH0R5VT436hRRRWpAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAHq3gL/kWI/8Arq/86KPAX/IsR/8AXV/50V81iP40vVnrUvgRwfi3/ka9R/66/wBBWLW14t/5GvUf+uv9BWLX0FD+FH0R5lT436hRVmz0+81GXyrK1muHGMiNC2M+vp+NLfaZfaZKI760lt2P3fMUgN9D3/Cr5le19SbO1yrRRRVCCiiigAoorQsdD1XUo/Ms9PuJo8E71jO049+mfapclFXY0m9jPoqSaCW2maGeJ4pV4ZHUqw+oNR0xBRU1naTX15Da26F5pXCKB6mtvxTp+l6LrcNlaLJL5Eafagz8O/UgccZHp69sVLmlJR6lKLtc56iun8WaFbWK2mq6Uv8AxKrxF2fMTsfHIOeecZ/P0rmKKc1OPMglFxdmFFd5qWneD9CNpb39rqUs0tskxaJwRzkdyOcg9qzvEel6NH4e07VtIiuYkupHTbOwJwCR057g96yjiIya0epbpNX12OUortND8JJD4vj0rWY4p0a2M2I5GA9uRg1yMLW5vEa5Rxbl/nWE4YL325z+tXGrGTaiS4NLUhoro/E3h+DT44NU0uYT6Tdn902eY2/un8j78EHpyvgnS7LV9cktr+Dzohbu4Xey4YEYPBHvS9tH2ftOgezfNynN0UV13hPRLMQPr+tNGNNgyFjYBvNbpyPQenc1VSoqceZijFydkcjRXf6j4Y8MxzWtxJqc9vHqI8y2SOAlcHHA6kdR1qeHw1oEdraeHr6dU1qbEzSxoSVyCdu7pjA7/XHIrH63C17P7v62NPYSvY85orf1/TdD0+No9P1Ke4vI5jHJFJFtCgZyc4weQPzq54X8IzapPZ3twsEmnyOQ6Gba5AJHQc9RWjrQUOd6IlU5OXKjlKK7bVvAN6dVuTp4tIrTf+6R7jkD3zk1d/4QL/ilMeTb/wBrebjzvtB27c/XHT2qHi6Vk77lewnd6HnlFWdQsZdNvpbOcoZYiAxRsjpnr+Nb+jaTp6eFNS1rVLd5VVhFaqHKEv69eRkj16GtZVIxipd/1M1Bt2OXoro9F0SW28WaZZ6raLsuMP5bkMGQg4PB9v0rL1W3C6/f29tEdq3MipGgzgBjwB7ChVE5cq7XBxaVyhRRXSeJdLsrDSdBntYfLkurXzJjuY7mwvPJ46npTlNKSj3BRbTfY5uiuubw28/hW2WPSL+LWEuCrhraT96p75xtUDI646Gl8W+HmsWtbew0q4ZbW2H2q8SF9kjYyT0xwOc5789KzWIg5cvr+BTpSSuchRVmHTr65j8yCzuJYycbo4mYfmBXS+C/D73PiFE1TTJjb+WxxNEyrnHFXUqxhFyfQUYOTSORorUvNE1BL24SHTbvylkYJiFiMZ45xVG4tLm0Ki5t5oS33fMQrn86pSi9mS00Q0UV23h/w9HrXgW8MMVsL77ZtSeXjaoCEjP5/nU1Kipq7KhBydkcTRXW/wDCvNX/AOfnT/8Av/8A/WrDtLSzi1z7Jq07xW0buk0kHzEEA4xwc8gdqUa0JJ8rvYHCS3M6ivQIvCvhObR5NWXUdS+xRttaRgq88dAVyeuOKxNR0nQHa0tdEvbuW+uJ44wlym1Qrjg/dHcr+BqI4iMnZJ/cU6TSOaorsfDWgxw+JNU0/VIIbhrWyd9ucqGBTBH4GuOrSNRSbS8vxIcWldhRXeaH4OW88GXF3PBCLu4YNbyTyNGI0HG449ck46HiqMfw71SV9kd9pjseyzkn/wBBrP61Su03sX7Gdk7bnI0VYv7NtPvprR5YpXhbazRHKk98Gq9bp3V0ZvQKK7Lwdolj9ml1vXQg05T5MQlBw7sQN30GcZ9cnjbWJ4k0SfQ9XlgkjCwOxe3dSSrJnjBPp0P+TWUa0XNwRTptR5jIoorovDOj2l/eNb6pb3kaXCbbedEIRGPIZj6enbmrnNQjzMUYuTsjnaK7Ow8KJFa+I4r22M89nGPs0qFgC3zcrjr245qGTw5a6f4VWe6guLnVbwB4I4AcQL/tcd+4PPYYwTWf1iF7L+upXspWuclRT2hlSXynjdZM42FSDn6VP/Zt/wD8+Vz/AN+m/wAK2uiLMq0V2uvaMieDdEmttO23b/64xxfOeP4u9cjJY3cMZkltZ0QdWaMgCs6dWM1dFSg4uxBRRVrTXtI9St3v4TNaBx5qBiCV78jn3rRuyuSirRXSeMNBTSdRW5syH028Hm27rjaM87RjsMjHsfY1zdTCanFSQ5RcXZhRWnpvh/VdXheaws3njRtrMGAwevc11Vx4Hni8JWvlaU8usySHzGWX/Vrk9RnGcYH4monXpwaTZUacpK6RwVFdB/whPiP/AKBcn/fa/wCNZupaNqGjvGuoWzQNICUBIOcfQ1UasJOyaJcJJXaKNFdn4O8O6bq2l3t1fIzvFKqKPO8tQCPWtz/hENB/59k/8GI/wrGeLhCTi+hpGjKSueYUV6NqvhHRodA1C7ghKTW8W9Cl2JOc9xivPI4pJpFjijaSRjhVUZJ/CtKVaNVNroROm4OzGUV1p8NiHwSt0+nXMmqT3JVAqPujQeq+nB5x/FUPg7w+mqa/Jb6jbTGC3haSWMhlOegHHOec474o9vDlcuw/Zyul3OYorWudB1U3Ezw6JqMcBYlFa3c7VzwCcdhVvwbpVlrGsS29+jtElu8gCNtOQR/TNVKrFRcuwlBt2OeortI7DwzqmhatdafZXkU1lCHBmlyCTnHT6VZ1jT/CGg3MNtd2WoSSyQrLmKQYwcjuR6Vl9ZV+WzuV7J2vdHBUV0d/ceEXsZlsbLUo7or+7eR1Kg+/J4qpZados9lHLd699lnbO6H7G77eePmHByOfxrRVdLtNfL/Ink1smY9Fd3pWjaDo0iajqV295YyxHZ5umyhOTwwPIzx+tVLjwhpsOnrqTa1NHYyN+7kfT3PB6Z5/XABqPrML21+5leylY4+itHVLPTbQQ/2fqv28vu8wfZ2i8vGMfe655/Ks6totSV0ZtWdgoooqhBRRRQB6t4C/5FiP/rq/86KPAX/IsR/9dX/nRXzWI/jS9WetS+BHB+Lf+Rr1H/rr/QVi1teLf+Rr1H/rr/QVi19BQ/hR9EeZU+N+p6H4cS5ufh/c2+gTCPVBOWuArYkK9sHPHGMH2PeuUvNT1W5SLStYup1iimDH7QhMkWepOfmPBzg+2Kr2y6rpccOqWy3NvG/3LhAQpw2CM9Oo6Gux1i5k1n4cpqWrwIl9HMEt5iArSqccge43cf7Oa52vZzvo0380zW/NG21l8izrWk6E3hHRkl1dLaBM+VdCyZjLnk/KOR681yWh+HRqVncaje3iWWm252yTMu4s3HyqPXkfmOtbfiEGX4eeHTH8/wA2z5eecEY+uQavaXdwxfDKOT+zo9RW3uG8+FmI28k7jgdgR+HNZxlOFPR3vK3Tuymoynquhz2peGLYaRJq+jaiL6yibbKrRlJI+g6dxz7fjT7Hwel1oFvrM+qRWtq5fzjJHnywpIGMHLEkdOKmfxVGdG1G203w9DaxzxhJ5Y2ZgozgZ49zj3NWru1uLn4T6a0MTSLDcvJJtGdq5cZ+nIrRzqxSUnbW3Ta33E8sG21roZmr+GLaDR/7Y0fURf2Cv5chKFHjPA5B9z6DqOvWqo8Va6Y7W3gvpYo7dUSKOD5R8oAGcfezjvmt7Qs2vw01uacFIp32RZB+Y8Dj8f5H0pnhzRE0nTF8S6jbSXDj5rKzVTuduzH27/Tnnij2iSan71nZeYcruuXS61JfiYYzd6WWCi8NuTPjrjjGfx3Vz174eez8N2Wsm6ikS6fYI1HK9e/4c+lVtautRv8AUHvdSSRZZjxuQqAB2UHsKgl068gsob2W2lS2mJEcpX5WP1rWlBwpxjf+uxE5KUm7HdeEdLtdLsYNR+02j6vfo4sVkbKRYUkliOh4wfy7ms+f4favO73c+qaWxmcs0hnbDMTk87frXGojyyLHGjO7kKqqMkk9ABXZeMNmk+H9G8PZBnhQzzgHO1jn+pb9KylGcai5Zay8uhalFw1WiNrRtHvNF0i8t9XudOvtC2lpo45WZoz6r8v6ZHqOevF+JdO0rT79RpOoLdwSrv2jnyh2BbuevuO9bGlkw/C/WpFJUyXSR5HplMj8ia5O0uZLO8guocebDIsiZGRkHI/lToQlzylfr8noKpJcqVj0Cwv9H8WRXFxqOijfptkCZDdv8yLk4woHv69a5nWvEVrqWj2emWel/Ybe1dnQfaDL1yT1GepPeuy0nxXrEvhbVNZv2iVIgEtSI8bpOn4jJX9az7jXr3Xvhzq1xfMhdLiONdi7Rjcp/rWELxnrHRNLd6f1c0lZx31t2Nz/AJqlb/8AYO/xrzbRYtDlM39tXF1CBt8r7OoOeuc5B9q9Bs721v8A4mwS2k8c8YsNpaNsjIzxXlNaYaLacdtF+pNZ9fN/odrqeiaDYzaalzqeoLplxam4iLYYqzFcYUDjI6/QVseDrXw1DrTtpWo3c9x5DgpLHgbeMnoPao9W0FdU0HR7y5u0tLK00pC8rDOWKjaAO/P49h1rF+G//Iyy/wDXrJ/Nazfv0JPmd0UvdqLTcY1h4JCMV1jUC2OB5PX/AMdrE0FVl1/TYJFDwyXcQeNhlWG4Dkd+CfzrOrpdA8O6vIbHWrK2iuI4rhXEfnqjHY2f4unSuySVOD5pb9zFXlJWR2+pWcK38ka2UHlodqD/AIR55gq9gHBwR9KSysbee/hSewt3QkAhvDrRDGP77EhasPbX93p91fz6VdR33mAJax6wyq44ycqwVe/HtS6XbXyxXN1NpV1DcW4DW8T6u0qzHng/Ngduvr7V5XNaO/5HZbXb8zzPxXDFb+KdQhgiSKJJcKiKFUDA6AVJ4O/5G7TP+u39DV7xNoWsSSX2u3tpFaxvIC0Xnq7DOAMY681c0NfClnqtpc215qc11GwKxCDduOOmAM16LqL2FlrpbTXocqi/aXempsa5paT61dynwZcXhZ8/aFvWQSe+McUzXrcW3w2aIaY+mgXIP2d5TIRz1yfWjXbvTIdZuBea5r1lMx3GCMkIoP8Adxxip7+XSj4Phtr+51h7Kd/NS7liLMeeMk9B6Z6jpXGnJKF77rv/AJ2N3b3v+AeaWVpJf3sNpEUWSZwil22qCfU16h4g0GC78KWel6bc20UVnOQ8077UyisHJIHXJOfxrzrWItIimjXSLi5njK5dp1C4PoMVvagxb4XaYzEljfOSSeT9+uuupScJJ21MKbSUkzrr2HR4/F2htcXFw+oiCNII4gPLK/NhmJ7deh9KqxazFdr4rtbbT4LT7Lb3BMsf35H+YFiePTNUtWmit/H3hyaeVIoksoizyMFUDL9SaoTeIlkm1HSNI0qCS9v5ZYGukf8A1qs7YwOn3T1zjvXJGk2l10Xy1N3NJv8Aroc/4d1bUNNvvJ05YWmu2WHEqbhknj9TXo3jx9Vt9HP2KGKS08kx3R2DcoPGR6DrXEeHJbXw5qd9c6rG6ahZRn7PAwwGkPHP4Ee2CT2rotY1u8t7Lw+puY0TUrLyrqSaMOoDbcuV6HGSfTk1tXXNWjKK/wCDp/kZ03am02WfDWkeJbfXLS5v9Z+0WgDF4vtjvuyhx8p68kGsfXNI8Tw29/dS60Ws/nLQi9c5Qn7u3p0PSrnhTQdJsvE1pcWviO3vJU37YEi2l8ow65PQHP4VlX/hvRWu7mU+KbVXLsxjMJyDk8feqYyXtb36L7L8/wCrlNPk2/ExLDxPrOl2gtbK+aGAEsFCKeT15IzXqXhO/vzptl/at0Zru/LSxqwAKRAcHj14P/Aq4Tw54Zi+zLr2tusGlxfOEccz+gA9M4+v612Wl2GoXHjQ6xd3ViYTC0UMEMuWRP4RjH5+5pYt05JpW9fPt/mFBTVmzK8Y67rNotnqWmX7R2VwpjZVVSElUnI5HsfyNcLqet6jrLRNqFyZzECEJUDGevQe1d/b2X9l22o22uXVncaNdTZ2xzbngZm4YDHHUZ54x9a4zxJ4dOgzxNHdw3NpcDfbyI4LFexI/qODWuFdNPlsr9H3/wCCRWU3qaXh3StFk8MX+r6vDcSLbTKmIGwcHaBxkDq1X4/EXg6LRJtIWx1X7JLL5rj5M7uO+/8A2RUeiyx6X4PaDVtLuriy1OcGMwsAWIxgYyGzlM9OeKXyPC/9qLpn/CN6p9tJAEPm/NyM/wB7059qmVpSlzXeuln2+fca0StYI9M8LatoWrXemWt9HLYw78zuMZIOOhOfumuT0e2s7vVreC/uDb2rsRJKCBtGD68dcV2VrfaZ9l1bRNC0PUFvLmJo5VZg20rkc5bjBJ/OuJ065gs9QhuLm1W7hQktAzbQ/Hrg1tS5rSWvlffYidrxPR9Q07Sdaslg0/XUt9K0mMSvFBAWI6/MWzy3Ddie/eotRvdC1nUNCvrXUVk1CC9hhI8oq0y715Ixxjrnp1FR+GdQ0ibRPED22hi2ijtszR/anfzlw/GT93vyPWsSOS1GoaTcw+HJdKt4L2OSW5eSR127l6lhwOPWuWMHzNO+np1XXvc1clZPTX1N3Tv+Sh+Jf+vOX+aV59p8trBqNvLewtNbJIGkjUgFh6c/y7+o613ukTw3PjzxFPbypLE9lKVdDkEZSuR0TxDNoQlNvaWc0jkbZJ4tzJ64PB5479q6KV9Ul0j5dDOdtL92dfpGo67rfiGS6/s5jok0YheGb5Ylh9QTwT1PGeuOmMSS6QumaRdN4PVb+a5Zo5ruKdWkgTg7FHuO/XofTD9d1rTTouky61bXV2b23EjJFO0aZGM/KCB1NZWkap4WbVbaCz0i/tpp5VhEsd46kbiB1DZxXOlJrmUbLt0076o1ur2b/wA9fkcdcWN3Z/8AHzazwZOP3sZX+ddPo3hvSYtMj1vXdSi+xt9y3hYl3P8AdOOc+oH1yKXxl4h1Nr7UNEa43WSzDAKjdgYIGepGefWuPrtXPVgm3a/bt+hzvlhLud/pl/b+NPE8en3Vr5ekwQt9mtUYoFxjBO3HOPwH887RfE1lPY/2L4jg87TgcwSDJa39AD1wO3cdOR0Phv8A8jan/XB/6VyNZqjFzlT6JK3lvqU5tRUvX9DofEug6dpCQT6fqyXkNwT5ajBYAdSWHHXjt+ldBo1x40vraOeTUvsGmqmftM8UYUKOmARk8d+nvXn1do+veFZ7KC1uNP1WSKFcJGbk7V69Bu9zTqwlyqLXM+9kwhJcze3zOrHisXmla0NOnd/7Othsu3AzLIQ3zYxjHy+n4VUi1u88TWYbQNYe01GOMGSyuFQhvUq23/PHAqxoVt4eTSGZLOaxh1Y/ZkS5lOZhg4xycA8gHjPHqM83PceFtHvzHLoerWt1Ef8AnuVYe4Ibp7iuKEINtRi7rbRP1v8AM6HKVk2/zOeF1ezeK4p9QbfeLdIJSQB8ysBjjjt2rp/F/izXNM8U3lnZ3zRW8ezagjQ4yik9RnqTXL3N5pceu291ptrNFZxOjmKRtzEg5Pc9a6DUPiPey30j2dparbnGwTxbn6DOTn1zXZODlKLULq3X5epzxkkmua2pr674k1e08G6LfQXhS5uP9a+xTu49CMVDpus6hrPgLX5dQuDO8a7VJUDAx7AVc1nxRfWXhLR9RiitTNc53q8WVHGeBniuel8ezXmhX9heWcXm3ChY3gGxQO+RXNTpSlDSC33+ZtKaUtZdP0MnwnoUXiHWxZzySRwrE0jmPG7jgdfcitDR9I0y70rxNK0Xn/YV3WsxdgQPnwTggH7oPSjwdqllo9nrNzLPGl81sVtkYHLHBJ56ddvGc8VN4Q/5FXxX/wBeyfykrpqyneT2S5fz1MoKOnzORluJ51jWaaSRY12oHYkKPQegrovEfhy2stPstW0mSS4024QBnbBKP7/Xn6EEelc7BC9xcRwR/fkcIv1JwK35tQ1fw9pV34cvbWMLOQ+2XD7Ae64OOoBB7EVtU5lKPI/l5GcbWfMaOjXE1r8NdYmt5pIpVuo8PGxUjlO4qxDpGqHTLO9uvGJtFu4/MjWe5dSemerc4yKpab/yS/Wv+vqP+aVPfXWga1oGi2txrTWc1jAUdfsjyZJC9xjpt/WuV35nb+bXS/RGytZX7d7dSPW7PWNI0mHUY/E815BNL5StDcORnB759sUvjGRZbXwxJdvK6vYxtKwOXIIXJ56nr1qLXNQ0lfBtjo2n6gbyW3uTIzGBo8g7z0P+8O9HjP8A5BXhn/sGp/6CtVBPmjfu+lhStZ28jY8Ox2zWE40L+3DamQeaHitGBYD/AGz9K1vst7gnytVwOv8Ao1h/jWH4OdV8LXdrNZzyCecOjNp8lzEQMdQuAeR68V0NpbQro+ooLKFVbysqNDkQN83eMtmTHt93rXLWfLN+vY2hrFFO+hnTS7oXY1lbMxn7Rsgslyo/3TmuI01rBPGOlnSmuxb/AGqEZuNofJYA/d4xXbTmO20DVoIrCQSXFuVX7NostuM4P3ic56+2Oa898Pf8jNpX/X5D/wChiujDq8Jv/gdDKr8UTu73UfFKeNfIhF7/AGb9rReLfKbMjPzY6dec1Pf3Uou9XstAZn124nMk5A2mKJAMYJ4OeBj/AGjVO+1HxKvjjyIXvf7P+2IuFjJTZkZ5x061iapqj6N8SLm/TJEVx86j+JCAGH5VlCnzWsltf19SpTtfff8AqxpeDte1bVb6/hvb6WaJLCVwrY65UZ/Wue8Kahc6XqM93b6ZLfgQMkiR5+RSRliQDgcVt6LdQ3Or+KdW09GggOnzMiycsGbBz3HVT+YrK8If8x7/ALA9x/7LW9opT0000M7v3de5v6k9ppfhmNdJ0J1Gs2u6YpM7+Vjp1znqfSp/F934fg1O1XVNNuLm4+yIQ8c20BcnAxn1zTL+91+28OeH49FE5SSz/feVCHHQYySDjvR4y1DRrXVLWPUNC+3TG0QiX7W8WBk8YA+vPvWEE3Jdd9nr+JrJ+6/l00MXX9O0YeFbDV9LtJbc3E7IVklLHA3D+YqPwTGbifWLbzYo/P02WJfNcKu8kBf681X1nxHa6jolrpVlpX2G3t5TIo+0GXrnI5Gep9at6NaaTF4OutWv9MF7NFdiJQZ3j+Uqv90+57V0vmVJqV9Xp1e+nX9TFWc7o7LT7W80jS1t9NkguJzFiQ3mol448f3VAxj8u1Q6hpza0lrdX5tI9St5EKPBffuioIJ+U9PoPQc0yFIPDWlrd23haZNRuwUNtDLJKyR56l+dueOBz09OMq40qwtvD41eTwdtUORJC97Oska9mII6fT298cUVeXN576f/ACR0Pa39fkc/43ZH8Yag0bKyEpgqcj7i1z9dN4ssNPtbfSLnT7P7Kt5aiZ4/MZ8E47sfeqt74U1Gx0KHWJjB9llVGUK5LYYZHGK9GlUiqcVfy+45Zxk5NmHRRRXQZBRRRQB6t4C/5FiP/rq/86KPAX/IsR/9dX/nRXzWI/jS9WetS+BHB+Lf+Rr1H/rr/QVi1teLf+Rr1H/rr/QVi19BQ/hR9EeZU+N+pt6R4t1jRIPItLkeRkkRSKGUE+ncVX1fX9T1yRWv7kyKpJRAAqr9AP8A9dZlFNUoKXNbUXPK1r6G3pXizWNGsmtLO5AhJJVXQNsJ7rnp/Kqmk63qGh3DTWFwYi4AdcAq4HqD/wDr5NZ9FP2cNdN9w55aa7G7qvi7WNYtWtbmdBbvjdFHGFDEHOT37Dv2rfTW7jSPh3pEthdLHcrdOGXIJK5fgj06VwdFZyoQaUUrJO5SqSTbNXWPEep67sF9cbo0OUiRQqqcYzgdT9fU1dj8deJIokjTUsIihVHkR8Af8BrnaKv2NNq3KrC55Xvc1dV8SavrUCQ6hd+dGjb1HlouDjH8IFQ3Gtahd6ZBps9xvtLc5ij2KNp5HXGT1PU1QopqnBKyWwnKT3Z02h+I9O0LTjJDpSy6xuIW4lbKqvOCB2POMDGfXtWL9ve41ZL6/JumMqyTB/4wDyPpjjFU6KSpxTcurBzbSXY6fX/FUGoaYmlaXpyWNgsnmEZyzn+nP17VgWMlrFfQvewNPbBv3kattLD2NV6KIU4wjyxCU3J3Z0XiPxQdZaC1t7VbfTLY/ubb1xwC2OnHYdMnnvU+r+JdNk0NtH0XSzaW0rrJK8jlmJHbqfQc5/CuWoqVQgkkuhXtJa+Z1vhvxHougWYuP7Nnl1ZVdfMD4RgTx349OnauSooqo01GTkt2S5NpLsbGt+Jb/XIraGcrHb26BUhjGFyBjcff+Q/HMvhTW4NA1aS7uIpJFaBowI8ZySPX6VhUUOlDk5LaD55c3N1CtHRpNIivC2sW9zNAF+VYHAyffOOPoR+NZ1FVJXViU7O5va/4k/taCGxtLSOy023YtHAnc8/MT68n8z1pdC1y20zRdbsp0maS+hVIigBAIDfeyf8AaHTNYFFR7KHJydCueXNzBXTaX4ls9D0cjTbArrEgKyXcpDBR6qO30x6ZzXM0VU4RmrSFGTi7olubma8uJLi4laWaQ5Z3OSTW34f8V3WjD7JOou9LfIltZACMHrtz0+nQ8/WufoolTjKPK1oCk07otajNbXGozy2dube2dyY4i2dorTudbt5vBlnoyxyi4guTKzkDYQd3TnOeR2rCoocE7X6ApNX8zrV8aQTW1umpaDZ308MYiE8hwSo6djUkHjawtJlntfDFjDOn3JFblT+VcdRWf1an2/Fle1mT3l3Nf3s13O26WZy7H3P9K1te1u31TTdHtoY5VaxtvKkLgYY4Xpg9OKwqK0cFdPsTzOz8zW8M6pDoviG11C4SR4od+5YwNxyhXv8AWqF5KlxfXE0YYJJIzqG64JzzUFFPlXNzdQ5naxKbmdrdbdppDArbliLHaD6gdM1r+E9at9B1tb65jlkjEbJiMAnJ+pFYdFKUFKLi9mCk07omu5luL2eZQQskjOAeuCc1HvyUDlmVeAM9BnOB6dTTaKq2lhHQ+JfEi6u9jHYxSWtpZxBYo88q3rx9Binaz4qbUdZsdVtrc293bIgZy+fMYc9BjAzkdeRXOUVmqMEkrbfqU6knc6W88VL/AMJeuu6dbNB08yNn/wBbxhs46ZH16ZqOXxBZJ4uk1i20xHt2YsbefaQWI+90O0554z9a56ihUYL7rfIPaS/U7KT4kaqCwtLSxtUPQJESR+OcfpWHqXibWdXt/s99fPLDuDbNqqCR9AKyaKUaFKLuog6k3uzsNN8S6Ho2jSCx0ub+1prbyZZXb5CT1PU8d8ACuPooqoU1BtrqKUnKyOwuPGVsuk6XaW+k2d09tAI5Wv7cPhsD7mG6df0qKz8aJDfW8snh/RURJFZmgs8SKAeSp3cN6e9cpRUfV6drWK9rK5o69qEWq65d30Kuscz7lV8ZHHfFZ1FFaxSikkQ3d3N3wlrVtoOuC+u0leMRsuIlBOT9SKxJNnmt5e7y8nbu647Z96bRSUEpOXcfM7WCuttdZ8LaXbxS2uizXV+qLlrqTKBsDP1wf9kVyVFKdNT0YRk47GjrOuX2u3YuL2QHaMJGowiD0ArcsfGUc9olj4isE1K2UbVl6TJ+Pf65B9zXJUUpUYOKjbYaqSTuW9TkspdRmfToZIbMn91HIcsBjvye/vVSiitErKxLd2dHrGv2uoeGNK0uGKZZbMfOzgbTx2wa5yiiphBQVkOUnJ3YVv8AhrxKmgR30M1gl7BeKqyRs+3pn2P941gUUThGceWWwRk4u6Oyh8X6FbzJNF4Rt1kRgyt9o6EdD9yub1nVJdZ1e41CVFR5mB2r0UAAAfkBVGiphRhB8y39WOVSUlZm3a63Fb+Er/RjC5kuZlkWQEYABX/CsSiirjFRvbqS23uaehXtjYaj5uo2IvLYoVMZ6g9iPxH6mrPiXxAuu3Ft5NotrbWsXlQxhskL/wDqxxWHRU+zi58/UfO+XlOsg8VaVbQLDDol3HGvRE1aVVz9AKtR+OrKK2nt10e62Tbd+7VZC3ByMEjI/DrXE0VDw1N73+9/5lqrNf8ADI6//hMNOwR/Y99gjBH9sTcj8qw4NStLXxDDqNvp5jt4ZVkS188nBXH8ZGeoz0rMoqo0YRvbr5sl1JPc6yX4gay+rfao5njtPND/AGXKkbc8ru2559fesbVdVTVfEEupyWuI5ZFdoPMPIAAI3ADrjr71mUURo04O8VboDqSlo2dfdeJ9Gs9KvLHw/pMkH2xCk008hJ2+wyexPes7QfFE3h+2uEtrK1knm4E8qZZV7r7g4HGfzrBopKhDlcXrcftJXub154z8QXqOkmoyJG38MIEeB6ZAzj8a17rxvpV+0cl74YhuJUQRh3uecD/gFcVRQ8PTfS3pp+QKrNdTqLzxHodxZTww+FreCWRCqSifJQkcHG3tVODWLaLwZc6OUl+0y3QmDADYFAUdc5zwe1YdFNUYJWXruJ1JM3J/GHiC4i8t9VnVQAP3eEP5qAaNM8Q3CXRj1W8vrjT5gVuYllJZ1wcAEn1x3GelYdFP2ULWSsHPK97nQeKNbstXexi0+3lhtbOAQoJSNx/U+g71a1DxEJvBFrpL3ou7kyKSBFtFvEoAVM4G4+/Pf2NcrRU+xhZLsHtJXb7hRRRWxAUUUUAereAv+RYj/wCur/zoo8Bf8ixH/wBdX/nRXzWI/jS9WetS+BHB+Lf+Rr1H/rr/AEFYtbXi3/ka9R/66/0FYtfQUP4UfRHmVPjfqFFFFakBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQB0PhjSUuNYthqemXcthN8u9VdVBPRsgdPx757UzW/C99pur3FpbW1zcwK37uVIi2QfXA6jofpVvw/wCK9bS/0zT1vmFr50UXl+Wn3NwGM4z0rR8VeLtd0/xNfWlrftHBG4CJ5aHHyg9xXG5VlWsrbd3/AJG9qfJch8XeH0sV06z0/S5pLiO3U3NxFGxV2xjtxngk/UVzFhpN/ql61nZ2zSXCglkyFIA65ziu58a+J9Z0rV7aCyvTFG9okjLsU5YlsnkewrhbXVLuz1ZdTik/0pZDJuI6k9cj0OT+dPDuq6V9PIKqgpmzb6Pbf8INqOoTQMt9b3YhBYkFR8uQV6Z5PatnRvDWnWfh2WTX3tbe6vh/oy3TbTGo/iwCDnJ5GemPer2paz/aXgnUdWtrQ2Za9idCRney7Pm9DyMfhRZW1zb3c2teJ9T02S0vLfYUb5y6EZATA49eM59M81zyqTcXd21+fTRGihFNW10/pmOvgW0ksbi9h8RW0tvbqWleOIsFwM9j19q5rSdIudau2trUxCRULnzH2jAIHX8a6fxJN/xSVqNIiS30Jrgoo8xjLM4zywx0yCQMnt06Dia6qDqSi23/AMAyqcqaSR6HqPgHOg6eLOG3TUP+XmRrg7W+mTj8hUfh7wDKNVH9sR209oUPyx3Bzu7H5SDWm1ot34H0NW0STVdqZ8tJzFs465HWpvC2nJa60sg8KT6afLYfaHu2kH0wR3ridaoqclzd+3+d/wADoVOLktO39bHCa/4autF3TymAW8kxSNUl3EDkjP4Cp/BelWWsavcW17CZUFq7oA5XDArg8fU1i6l/yFbz/ru//oRruPA9/wCHYtSsoLe0vF1OaMxySuw8snbk8buhx6V11ZTjRfV+RzwUXU8jmtL0rT9R8PajKbpYtStiJEWVwqNH0IGe5Jx9dvrWd/ZV9/Zf9p/Z2+x7tnm5GM+nrXd2c8VzqGoWul+FtKjfTy++a5bcoKkgfw9TjPXseay9Una6+GtvcMqI0uqu5VBhQSHOB7Uo1pc1u7X4+hTpq3yM7xhpdnpN7Yw2cXlrLZpM+WJJYkg9T7Vm6HpUutaxb2MQP7xvnYfwoOp/Kt74hf8AIV03/sGxf+hPXQaLpMei6U9vp99Y/wBvXtt5vnyP8kUWQPlIB9ePUjPal7ZxoJ9WHs06j7I5DUY9EsfGLQRwySaXDMI5VaQ5OOGwRzgHPrnHXmjxboI0PVf3HNhcjzLZg2fl7jPtn8sVpP8ADjU41Vn1LSlVxlSZ2AI9vlrobTTW0/w41l4nuLG80hXVYZoZWZ4WJwOdo459eORyOkuvGLi4yv0a7/8ABGqbd01Y850uOwl1KJNTmlhsznzHiGWHBxjg98V2kXhXwnNo8mrLqOpfYo22tIwVeeOgK5PXHFc/JpGjxeKo9P8A7XEunMfnukwNvBIGeQewz712OoadpOtWSwafrqW+laTGJXiggLEdfmLZ5bhuxPfvTr1NYtNpPy/4G4U4aO6TOT1HSdAdrS10S9u5b64njjCXKbVCuOD90dyv4GrnhrQY4fEmqafqkENw1rZO+3OVDApgj8DW3qN7oWs6hoV9a6ismoQXsMJHlFWmXevJGOMdc9Oopunf8lD8S/8AXnL/ADSs3Vm4NO+3Xff5FckeZPzOW8G6VZaxrEtvfo7RJbvIAjbTkEf0zWnHYeGdU0LVrrT7K8imsoQ4M0uQSc46fSsnwpqFzpeoz3dvpkt+BAySJHn5FJGWJAOBxXS6k9ppfhmNdJ0J1Gs2u6YpM7+Vjp1znqfStKzkqllfW1tfv/AiCXJ95DrGn+ENBuYba7stQklkhWXMUgxg5Hcj0rFv7jwi9jMtjZalHdFf3byOpUH35PFdP4vu/D8Gp2q6pptxc3H2RCHjm2gLk4GM+uawtf07Rh4VsNX0u0ltzcTshWSUscDcP5ipoyuouXNr56FVFZu1tDOg0vQZLaKSbxH5UrIC8X2GRtjY5Gc4OPWuhtdG8P6Hb3SavdSSLdRBYZZdOkQxkg8qSCM8/pVHwpaHUfDeu2ImhRpWt/LEsoQcOS2OuOB1xXaL/aOn2XkaTJazgbWefUb8ykH6DgD6EfSpr1WpOHM9/Jdn2HTgmua35/5nFX3g/T9NghnvdblgjnGY2fT5Of14Psa57U7axtblEsNQ+3RFAzSeS0WGyflw3tg5969RfTYrvW9P1ic28V7FJ+/232+Mpgj5QRn8BgcnrXmupRx3Piq8iMyRRy3rqZWPyqC5+Y+w61rh6spv3nt6foiKsFFaI3fDPhSz1nw/NqEyXssyXJiEVvLGmRtU5+cY7nvWj/wg9n/0C9a/8DLX/GrGn6b/AGRoAgbWNCn064mMitdRF1ZwADj5gONtTT2NtapC88vhONZk3xlrMgOvqPm5FYSrTcnaWnzNFCNldfkc/wCJ/Ctno2hW99Ct7FNJP5TRXEsb4GGOfkGOw71ytotu95Ct3I8duXAkdF3Mq9yBXfapDbappkWnNr/h+3topPNVbZCmDgj+9jua5Cy0/TJb66gvtZW1iiYrHMtu0olwcZG3oO9dNCo+R8z19GZVI+9oWfE3h06JcRzW0hn025UNbXGQdwx0OO/8x+ONi98MeGNGuBbaprl0bjaGKxW5GQenY9vesjxJok+kajaaWt896jRB4cqUC72PABJxkjP411/iV/DFx4rFlqFrfS3p8uItEwCDIGO/oRUSqStFJtqz2+XcpRV3p95yWs/8IsmnrDo/257tXDGeYDay85U8jHY8L2rK07S7rUriOOCCVkaRUaRIywTJ6mus1qDwdo1/dafJp+oPcRDAcSfLkqCP4h6ijwRNeReHvEbWRk+0JHE0QQbju+foPXiqVVxpc0b9N/MnkvOz/AzfF2iJY63PHpmm3MdlCgDPtdlJxkncc8c469qsr4ZU+BIL6LTbq51K6mPlmJXbZHnqVHGDt64/irb0i81m80HXF8RPdxwCJFDywbSqkkOQMDJxVHxFrurQw299osz22hDFtbNGQN5UHnaRkdCOf7tZqpUdqaeqe999C3GCvLucddaXqFiiveWN1boxwGmhZAT6cir3h+wt7jUoTqVpePYyZXfChwCeMk+g749K2/EN5c3/AIA0S6u5mmmknl3O3U4LAfyp3hyfxle2cEWn3Jt9OjG0TSRoERR15IyfwzWsqsnTbdlut7EKCU7bkumeDYIPFN5YX0f2u0S1aWKQFlGcjbkjHOM8VR03w5a2/h241PVYLiaaYGK1tYQQ4b+8eOPx4x2ORXc6V4linvW0mG9N/JBavLNekKqs4IACgcY5/l1rI0vxJceItOhtLTWZLDWkQjbKiNHcEd87ePoOnocZrl9rXd7+V9/Py0ubclPp5nmksMsD7JonjYjOHUg4/Go61fEU2qyaxJHrMnmXkAEZOF6dR93jvn8ayq9OLbimzjas7BRRRVCCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigD1bwF/yLEf8A11f+dFHgL/kWI/8Arq/86K+axH8aXqz1qXwI4Pxb/wAjXqP/AF1/oKxa2vFv/I16j/11/oKxa+gofwo+iPMqfG/UKKKK1ICiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACpLed7a5injxvicOuRxkHIqOikB6N4Y8X+Itd163sy0Hk53zERdEHXv34H41T1b4i6rHq11HYSW5tUkKxExg5A4znPOaov4vgtPDq6doth9hmlBFzNu3E/7p68juenQetcnXHDDQlJylBJdF+p0SqySSUj07xf4w1bRdUt7e0aERvbJK2+PJ3EnP8AKuUgvovFPiSGbxBeR28QTDOq7QwXkL7Zyeah8V63Br2pw3NtFJEkdusW2TGcgk9vrWFVUKCjTWlpW+YqlVuW90enahrsF74JvbjT7dIrWyvIo7ZCvBVShBI9yaZH4k1S78Jy6xapYST28u24gFsf3adiOcnsc9MZ9K5K21q2h8F3ujMkpuJ7lZVYAbABt6nOc8HtWNFcSxI0au3lOQZItxCyYOQGAPNRHCqzVtn+Gg3Wd7+R6Tc+Lb+HwPY6qILQzzXTRshi+TA3dBnrwK46wGhXxubjWbu6t7iSYsEtowVwee+e+at+IfEOl3mi2ukaRZTW9rBKZczNk5OeOp/vHvWXoV1pdnf+dqtlJdxKuUjVsDd2yO4/zg1VKlywbSab+8U53kle539zdaYPCljsvdZtLC3by0uYYyhk44z7e+MZo8N3lh9tlubLVNd1IQRFpIpQXUA+3c+g69a4PXfEd/r8+bhwluh/dW6cJGP6n3NUtP1G80q7W6sbh4JgMbl7j0IPBHsahYRum03q/wCuxXt0peRuavD4XKXc1te6g16zFlieIKu4noeOBUPhLSdRvtZtruzQrDbTK8k7HaiAckE/Tt70eItfs9egtpv7OSDUh/x8zIcLJ2HHf8enTmqtx4k1S40e30prjZawrt2xjaXHYMR1xW0Y1PZ26vv0IbjzX/I7Lx5eXEdvBbaVAPsOpN5slxB8xuJD/Dx7Y+v4GsK7uLdPh9Fpk0pi1CC+8xrZ0IfBB7Hpw2c1H4c8XNo9v9hvbVb2w3+YqMfmjfIIK546jOPXmkXxDp134su9W1XT/Ptp1wIeCVwAFPJHZf1rKFKUFy20Wt+5cpqWt99PQsfEL/kK6b/2DYv/AEJ65mys59QvYbO2QvNM4VQB+p9h1NaXifXV8QaolzHbfZ4oohDGm7J2gkg/r0rQsfFVno2jLFpOmCLVJEKTXkh3Ee6/XrjgA+taw54UkktSJcsptt6Enj64gS/sdItzuj023WIn/aIHH5BakviYvhVpaqSPOvGZsdwC4/oPyrnNMvLeDWIrvUoTeQ7i0yMcl8g8898nNaPiTxMNbjtrS2s0s7C1z5USnJPufw7fXk1PspLkglotW/vHzp80n1MnTrmCz1CG4ubVbuFCS0DNtD8euDXfeGdQ0ibRPED22hi2ijtszR/anfzlw/GT93vyPWuG0bUk0nVIryS0iulTOYpQCDxweQcEHBz7V0cnxI1UFhaWljaoegSIkj8c4/SjEU5z0ivxt17BSlGOrf4EUclqNQ0m5h8OS6VbwXsckty8kjrt3L1LDgcetbOkTw3PjzxFPbypLE9lKVdDkEZSuP1LxNrOr2/2e+vnlh3Btm1VBI+gFbWm+JdD0bRpBY6XN/a01t5Msrt8hJ6nqeO+ABUVKU+XbV6b3/FjjON/L+uxT8If8x7/ALA9x/7LXR397r9t4c8Px6KJykln++8qEOOgxkkHHeuW0HxRN4ftrhLaytZJ5uBPKmWVe6+4OBxn86Lzxn4gvUdJNRkSNv4YQI8D0yBnH4050ZzqXsrefoEZxjC19TqfGWoaNa6pax6hoX26Y2iES/a3iwMnjAH15965nWfEdrqOiWulWWlfYbe3lMij7QZeucjkZ6n1rUuvG+lX7RyXvhiG4lRBGHe55wP+AVSvPEeh3FlPDD4Wt4JZEKpKJ8lCRwcbe1TSpyioqUHdeen3XHOSbdpL7v8AgFu0sdItvCuk3kuif2heXs7w4Fy8ZJ3EDoceg6V0N2Lfw/pYsLPwtPNNfRhryCCSRggxwpkAJJ68D39eeKm16L/hGdJsLYTJeWM7TGQqNuSxIxzz26iornxd4guxiXVbkDOf3beX/wCg4puhUm7va76vvoCqRitPyR0mq6dpukaVaahP4RUxTr84+3TBoW7KwI9P1yPTOF4002z0rX/s1jD5MPko+zcW5PXkkmk0vX0YyWuvzX15pzjcYklJJfjBJJHHXjPvio/EmvJrHiD+0bWNoVQIIw+Cfl7kdOtVThUjUs/Pq7eW5M5RcdDs9CuntvA2lbLpYGaSUc2bXG75z2XkfWtS8v5bHU9GmuFRrO7hFvJcNDs2yHJT5TyvOeO3euBufHniC44jukto8fcgiVR+Zyf1q3ceK9Ovk046hZ3V61pbBSksuEkmGPmbruHXr+ua55YafNzNb328/uNVWjayfY1PFF7qFr4Re11iO1F9dXOxFhULiJCDu/Ej8iK87rqh41kvvMi13T7fULZ2JVceW8WeyMOw/P3rO8P61b6LqUl5Lp8Vw2xvJBJ/dP2Iz27c8+/r1UYzpwa5dfLYxqOM5LU7jVdMW88eabPcypBa2tklxI8jhc7CxwM9ecZ9B+Fcdcakur+O0vkGI5b2PZ/uhgB+gFWYvHeqO0ialFa6jbSHJguIhhf9307dc9KlHiPwzcSq1z4VWIryrW1yy88dhtrKFOpT+KN9Laf8GxcpRls7alXx6MeNdQA9Yz/5DWtHwTJcw+HvEslmZBcrHCYzGMtn5+lNfxF4VCzPH4ele4deHuJjJzjj7xNZeheJ5NA02/gtrc/abrbtuBJjy8Zx8uDnqapxnKjyKO1t7dBXiqnNfudFYXmt3fhbxF/bDXRC26+V56Fe5zjj6Vz2ma/Bb+GNS0a9iklSfD2+0/cf1PPTIU8e/rViPxxqUml6hZagzXguovLRmYL5XXJwF56j8qzdCudFt5pv7asJbqNlHl+U5Uq35jg/0pxpNKXNHqnp8vQTmm1Z/eamvCSDwR4Ztyw2uJ5SB7sCP0Y1JZ654cTRINPvLLU5Ao3SolwRGz8ZON3tWd4k8QJrj2sVvaLa2dohjgjDZODjr+AHH6mrOmaj4WsNPhkuNJuLzUVUlhJJ+6zk4/8A1YP40cj9muZO929PNsOZc2j08zr/AAwfDdvbzazb2Nxp9uMW4nuZMq+4gYHJ6HGT0H4GsbVIvC2hag1vdaFqUciNlJFmO1x2ZTu5H+etc3rviS+150WfZFbRf6q3iGET8O5x3q9pHjCS1s003VbSLUtOXhUlHzxj/ZPt6fqKz+r1F7+rv0v+pXtIv3f0M/XrvSby5SXS7a5h3ZMxuJN7OxPXOTWTWjrU2lT3+/R7Wa2tdgGyVsnd3PU/z7VnV201aK/UwluFFFFWSFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAHq3gL/AJFiP/rq/wDOijwF/wAixH/11f8AnRXzWI/jS9WetS+BHB+Lf+Rr1H/rr/QVi1teLf8Aka9R/wCuv9BWLX0FD+FH0R5lT436hRRRWpAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAHq3gL/kWI/+ur/zoo8Bf8ixH/11f+dFfNYj+NL1Z61L4EcH4t/5GvUf+uv9BWLW14t/5GvUf+uv9BWLX0FD+FH0R5lT436hRRRWpAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAHq3gL/kWI/8Arq/86KPAX/IsR/8AXV/50V81iP40vVnrUvgRqX3gTRtQvZbu4WUyytuYhyKr/wDCuNB/55zf9/DXX0UlXqpWUn943Tg9bHIf8K40H/nnN/38NH/CuNB/55zf9/DXX0U/rFX+Z/eL2UOyOQ/4VxoP/POb/v4aP+FcaD/zzm/7+Guvoo+sVf5n94eyh2RyH/CuNB/55zf9/DR/wrjQf+ec3/fw119FH1ir/M/vD2UOyOQ/4VxoP/POb/v4aP8AhXGg/wDPOb/v4a6+ij6xV/mf3h7KHZHIf8K40H/nnN/38NH/AArjQf8AnnN/38NdfRR9Yq/zP7w9lDsjkP8AhXGg/wDPOb/v4aP+FcaD/wA85v8Av4a6+ij6xV/mf3h7KHZHIf8ACuNB/wCec3/fw0f8K40H/nnN/wB/DXX0UfWKv8z+8PZQ7I5D/hXGg/8APOb/AL+Gj/hXGg/885v+/hrr6KPrFX+Z/eHsodkch/wrjQf+ec3/AH8NH/CuNB/55zf9/DXX0UfWKv8AM/vD2UOyOQ/4VxoP/POb/v4aP+FcaD/zzm/7+Guvoo+sVf5n94eyh2RyH/CuNB/55zf9/DR/wrjQf+ec3/fw119FH1ir/M/vD2UOyOQ/4VxoP/POb/v4aP8AhXGg/wDPOb/v4a6+ij6xV/mf3h7KHZHIf8K40H/nnN/38NH/AArjQf8AnnN/38NdfRR9Yq/zP7w9lDsjkP8AhXGg/wDPOb/v4aP+FcaD/wA85v8Av4a6+ij6xV/mf3h7KHZHIf8ACuNB/wCec3/fw0f8K40H/nnN/wB/DXX0UfWKv8z+8PZQ7I5D/hXGg/8APOb/AL+Gj/hXGg/885v+/hrr6KPrFX+Z/eHsodkch/wrjQf+ec3/AH8NH/CuNB/55zf9/DXX0UfWKv8AM/vD2UOyOQ/4VxoP/POb/v4aP+FcaD/zzm/7+Guvoo+sVf5n94eyh2RyH/CuNB/55zf9/DR/wrjQf+ec3/fw119FH1ir/M/vD2UOyOQ/4VxoP/POb/v4aP8AhXGg/wDPOb/v4a6+ij6xV/mf3h7KHZHIf8K40H/nnN/38NH/AArjQf8AnnN/38NdfRR9Yq/zP7w9lDsjkP8AhXGg/wDPOb/v4aP+FcaD/wA85v8Av4a6+ij6xV/mf3h7KHZHnujeBtGvjqPnRyf6PeyQJiQ/dUDH860/+FcaD/zzm/7+GtLwz97Wv+wpN/Ja3a0q16qm0pP7yYU4OOxyH/CuNB/55zf9/DR/wrjQf+ec3/fw119FZ/WKv8z+8r2UOyOQ/wCFcaD/AM85v+/ho/4VxoP/ADzm/wC/hrr6KPrFX+Z/eHsodkch/wAK40H/AJ5zf9/DR/wrjQf+ec3/AH8NdfRR9Yq/zP7w9lDsjkP+FcaD/wA85v8Av4aP+FcaD/zzm/7+Guvoo+sVf5n94eyh2RyH/CuNB/55zf8Afw0f8K40H/nnN/38NdfRR9Yq/wAz+8PZQ7I5D/hXGg/885v+/ho/4VxoP/POb/v4a6+ij6xV/mf3h7KHZHIf8K40H/nnN/38NH/CuNB/55zf9/DXX0UfWKv8z+8PZQ7I5D/hXGg/885v+/ho/wCFcaD/AM85v+/hrr6KPrFX+Z/eHsodkch/wrjQf+ec3/fw0f8ACuNB/wCec3/fw119FH1ir/M/vD2UOyOQ/wCFcaD/AM85v+/ho/4VxoP/ADzm/wC/hrr6KPrFX+Z/eHsodkch/wAK40H/AJ5zf9/DR/wrjQf+ec3/AH8NdfRR9Yq/zP7w9lDsjkP+FcaD/wA85v8Av4aP+FcaD/zzm/7+Guvoo+sVf5n94eyh2RyH/CuNB/55zf8Afw0f8K40H/nnN/38NdfRR9Yq/wAz+8PZQ7I5D/hXGg/885v+/ho/4VxoP/POb/v4a6+ij6xV/mf3h7KHZHIf8K40H/nnN/38NH/CuNB/55zf9/DXX0UfWKv8z+8PZQ7I5D/hXGg/885v+/ho/wCFcaD/AM85v+/hrr6KPrFX+Z/eHsodkch/wrjQf+ec3/fw0f8ACuNB/wCec3/fw119FH1ir/M/vD2UOyOQ/wCFcaD/AM85v+/ho/4VxoP/ADzm/wC/hrr6KPrFX+Z/eHsodkch/wAK40H/AJ5zf9/DR/wrjQf+ec3/AH8NdfRR9Yq/zP7w9lDsjkP+FcaD/wA85v8Av4aP+FcaD/zzm/7+Guvoo+sVf5n94eyh2RyH/CuNB/55zf8Afw0f8K40H/nnN/38NdfRR9Yq/wAz+8PZQ7I5D/hXGg/885v+/ho/4VxoP/POb/v4a6+ij6xV/mf3h7KHZGXpmg2ek2YtbbeIgxYAtnrRWpRWTbbuy0klZBRRRSGFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFAGF4Z+9rX/YUm/ktbtYXhn72tf9hSb+S1u1pW+NkQ+EKKKKzLCiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAMLwz97Wv+wpN/Ja3awvDP3ta/wCwpN/Ja3a0rfGyIfCFFFFZlhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFZ2r63ZaHbpPfu6RO2wMqFueuOKcYuTshNpK7Kfhn72tf9hSb+S1u1wnh7xho0VxfwtPIZLvUHkhVYmJYNtC9vWu7rbEQlGfvKxFOScdAooorA0CiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAoryDxH4p1yy8Q3tvb6jLHDHJhVGOOB7Vlf8Jn4i/6Cs35L/hXdDL6soqSa1/rsc0sVBO1me6UV4X/AMJn4i/6Cs35L/hR/wAJn4i/6Cs35L/hVf2bV7r8f8hfW4dme6UV4X/wmfiL/oKzfkv+FH/CZ+Iv+grN+S/4Uf2bV7r8f8g+tw7M90orwv8A4TPxF/0FZvyX/Cj/AITPxF/0FZvyX/Cj+zavdfj/AJB9bh2Z7pRXhf8AwmfiL/oKzfkv+FH/AAmfiL/oKzfkv+FH9m1e6/H/ACD63Dsz3SivC/8AhM/EX/QVm/Jf8KP+Ez8Rf9BWb8l/wo/s2r3X4/5B9bh2Z7pRXhf/AAmfiL/oKzfkv+FH/CZ+Iv8AoKzfkv8AhR/ZtXuvx/yD63Dsz3SivC/+Ez8Rf9BWb8l/wo/4TPxF/wBBWb8l/wAKP7Nq91+P+QfW4dme6UV4X/wmfiL/AKCs35L/AIUf8Jn4i/6Cs35L/hR/ZtXuvx/yD63Dsz3SivC/+Ez8Rf8AQVm/Jf8ACj/hM/EX/QVm/Jf8KP7Nq91+P+QfW4dme6UV4X/wmfiL/oKzfkv+FH/CZ+Iv+grN+S/4Uf2bV7r8f8g+tw7M90orwv8A4TPxF/0FZvyX/Cj/AITPxF/0FZvyX/Cj+zavdfj/AJB9bh2Z7pRXhf8AwmfiL/oKzfkv+FH/AAmfiL/oKzfkv+FH9m1e6/H/ACD63Dsz3SivC/8AhM/EX/QVm/Jf8KP+Ez8Rf9BWb8l/wo/s2r3X4/5B9bh2Z7pRXhf/AAmfiL/oKzfkv+FH/CZ+Iv8AoKzfkv8AhR/ZtXuvx/yD63Dsz3SivC/+Ez8Rf9BWb8l/wo/4TPxF/wBBWb8l/wAKP7Nq91+P+QfW4dme6UV4X/wmfiL/AKCs35L/AIUf8Jn4i/6Cs35L/hR/ZtXuvx/yD63Dsz3SivC/+Ez8Rf8AQVm/Jf8ACj/hM/EX/QVm/Jf8KP7Nq91+P+QfW4dme6UV4X/wmfiL/oKzfkv+FH/CZ+Iv+grN+S/4Uf2bV7r8f8g+tw7M90orwv8A4TPxF/0FZvyX/Cj/AITPxF/0FZvyX/Cj+zavdfj/AJB9bh2Z7pRXhf8AwmfiL/oKzfkv+FH/AAmfiL/oKzfkv+FH9m1e6/H/ACD63Dsz3SivC/8AhM/EX/QVm/Jf8KP+Ez8Rf9BWb8l/wo/s2r3X4/5B9bh2Z7pWfrmlR61o9zYSYHmL8jH+FhyD+deN/wDCZ+Iv+grN+S/4Uf8ACZ+Iv+grN+S/4U45fWi7pr8f8geKptWaZ0Pw88OyNrdxe3kRUWDGJVYf8teh/IfzFep14Sni/X492zUpF3NubCqMn1PHWnf8Jn4i/wCgrN+S/wCFa18HWrT5m1+P+RFOvTpxsrnulFeF/wDCZ+Iv+grN+S/4Uf8ACZ+Iv+grN+S/4Vj/AGbV7r8f8i/rcOzPdKK8L/4TPxF/0FZvyX/Cj/hM/EX/AEFZvyX/AAo/s2r3X4/5B9bh2Z7pRXhf/CZ+Iv8AoKzfkv8AhR/wmfiL/oKzfkv+FH9m1e6/H/IPrcOzPdKK8L/4TPxF/wBBWb8l/wAKP+Ez8Rf9BWb8l/wo/s2r3X4/5B9bh2Z7pRXhf/CZ+Iv+grN+S/4Uf8Jn4i/6Cs35L/hR/ZtXuvx/yD63Dsz3SivC/wDhM/EX/QVm/Jf8KP8AhM/EX/QVm/Jf8KP7Nq91+P8AkH1uHZnulFeF/wDCZ+Iv+grN+S/4Uf8ACZ+Iv+grN+S/4Uf2bV7r8f8AIPrcOzPdKK8L/wCEz8Rf9BWb8l/wo/4TPxF/0FZvyX/Cj+zavdfj/kH1uHZnulFeF/8ACZ+Iv+grN+S/4Uf8Jn4i/wCgrN+S/wCFH9m1e6/H/IPrcOzPdKK8L/4TPxF/0FZvyX/Cj/hM/EX/AEFZvyX/AAo/s2r3X4/5B9bh2Z7pRXhf/CZ+Iv8AoKzfkv8AhR/wmfiL/oKzfkv+FH9m1e6/H/IPrcOzPdKK8L/4TPxF/wBBWb8l/wAKP+Ez8Rf9BWb8l/wo/s2r3X4/5B9bh2Z7pRXhf/CZ+Iv+grN+S/4Uf8Jn4i/6Cs35L/hR/ZtXuvx/yD63Dsz3SivC/wDhM/EX/QVm/Jf8KP8AhM/EX/QVm/Jf8KP7Nq91+P8AkH1uHZnulFeF/wDCZ+Iv+grN+S/4Uf8ACZ+Iv+grN+S/4Uf2bV7r8f8AIPrcOzPdKK8L/wCEz8Rf9BWb8l/wo/4TPxF/0FZvyX/Cj+zavdfj/kH1uHZnulFeF/8ACZ+Iv+grN+S/4Uf8Jn4i/wCgrN+S/wCFH9m1e6/H/IPrcOzPdKK8L/4TPxF/0FZvyX/Cj/hM/EX/AEFZvyX/AAo/s2r3X4/5B9bh2Z7pRXhf/CZ+Iv8AoKzfkv8AhR/wmfiL/oKzfkv+FH9m1e6/H/IPrcOzPdKK8L/4TPxF/wBBWb8l/wAKP+Ez8Rf9BWb8l/wo/s2r3X4/5B9bh2Z7pRXhf/CZ+Iv+grN+S/4Uf8Jn4i/6Cs35L/hR/ZtXuvx/yD63Dsz3SivC/wDhM/EX/QVm/Jf8KP8AhM/EX/QVm/Jf8KP7Nq91+P8AkH1uHZnulFcf4P1S91DQFuLu5eWUyMCzegorhnBwk4vodMZcyujzjxb/AMjXqP8A11/oKxa2vFv/ACNeo/8AXX+grFr6Sh/Cj6I8mp8b9QooorUgKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigD1bwF/yLEf/XV/50UeAv8AkWI/+ur/AM6K+axH8aXqz1qXwI4Pxb/yNeo/9df6CsWtrxb/AMjXqP8A11/oKxa+gofwo+iPMqfG/UKKKK1ICiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooA9W8Bf8ixH/wBdX/nRR4C/5FiP/rq/86K+axH8aXqz1qXwI4Pxb/yNeo/9df6CsWtrxb/yNeo/9df6CsWvoKH8KPojzKnxv1CiiitSAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAPVvAX/IsR/wDXV/50UeAv+RYj/wCur/zor5rEfxperPWpfAjg/Fv/ACNeo/8AXX+grFra8W/8jXqP/XX+grFr6Ch/Cj6I8yp8b9QooorUgKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAK0rbw/rF5bG4t9NupIQNwcRnDD29fwqnaeR9sg+0gm38xfNx/dzz09q9D8ayeIILqHUtKupjpIjVojaMdqDGcsB1HHU8Y4+uFWq4yUV17mkIJpt9Dzh0eN2R1KupIZWGCCOxptdb4Xki1vxhFe6reKLoyIyIYNwnYDHbhegOaZ42sdMt9Xu5rXUVe6e4/eWa2zIIwRknd0POOnrQq37z2bWtv6/4cPZ+7zI5WiuuPhPTdOihj13WhZ3s6BlgSIv5eem8j/63Q81m6h4YuNN8QWulzzIVunQRToMqys2Acevtn8aca8JOyYOnJGHTkRpHVEUszHCqBkk+ldhceDdL0u8Ntq/iGO3kkJECrCSSOgZ+cKM9ienfrWXdWWpeDvEqxxMr3Sgm3kVd28MCoIHryePUd6I14T0juDpyjuU5vD2sW9p9qm0y6jhAJLNERtA7kdQPc1m16T4NuPEK6rdXOtSXiaeImaZr3KoD2wG6d+nb8K4eG0i1XxCLS1dIYbi4KxM/AVS3H6dqmnWblJStp1Q5QSSa6jNJh06e92apdSW1ttJ8xELHPYYwa3P7O8G/wDQdvf/AAGP+FYmtaY2javcae8qytCQN6jAOQD/AFr1XxBeSwX0SJ4qh0oeSp8h7dXJ5PzZP5fhWVeo04uLev8AXZsunHR3W39d0cjfeGfDGmw2ktzrN4qXUQmiIgzlT9Bx1rMvLDwrHZTPaaxdS3CrmON4CAx9M4r0DV7yWK204r4thsi9urGRrdW+0cD5xnpn0965f4huZLDQXa6F2WikJuAu0Sfc+bHbNY0Ks5Sim3r/AF/L+pdSEUm0v6+8wdE8P22r2kksms2lnIrkeVMeSoXJbr06/ka6bU/DGh6hbWkljrGl2v2aFY7p43yrN2brxk56nJ/Cs3wjdaFpPlahcatJDelXjltzbs6FT0zgfQ8HtW5LZ+HND8PvZNqs0EerBZxMsLFmi4IXgdOe/qaqtUmqmjflp9/QUIx5dbfeYUWg6DY6ENV1Ka9uU+2NbL9kZNr4JweexC9j3rP1e98NT6eItK0y5trhW3CWSTduHcHk11J/4R218DRLdm7vNMGoMIXhG12bDYJzjjG79KyfFum+G9Htkt7S3vBfTwpPE7OCgUt0PPXAPaqp1Lz9697/ACFONo6W2OUsbObUL6Gzt13TTOEUfXufYdTXe+IvAatc266Y1haosIDma4ZWkbpnGCB07dTmub8N60ukQXgtLKSXVrhRHbyqA3ljvhcZz37546Y56m00ttS8NQweLmFiIpB9kuppVWUqeSpDf1/Ljl4ipOM072S/H5CpRi423Zzd/wCB9Q0/TJtQmvNPa3iHJSYnJ9B8vWuZrt/F+na0xhtLXSpYtHtflt47c+YrdfnOOckevv6knkrWwludRjsmZLeWRtuZyUCntn0rahUcoc0ncipFKVkivHG8sixxozyOQqqoyWJ6AD1r0SbwfZL4ebSIRBJ4jjjF03JDEE8qDnHQY/XjOart/YvgRW8qRNR18KQpx8kGf6/r9AaozMYfClr4njaQa1JflWui5z0Y9M46ADGMY4rGpUlUacdFfTzf+RcYqN09X/X4nIujRuyOpV1OGVhgg+hojjkmkEcSM7noqjJNd6s3h7xuqm9kXS9bICtIOElPQdeD1HGQe2TiuZtft+keI5LfRLkz3Ss0KSQxZ3+uAQfTr7VvGs5Jq1mu/wDmZuFrO90af/CM2174V+22iXMGoWv/AB8Q3Ckebnun8h9Pxqzf+EknXw7BY2xt5ryAG5kcsQGwCScng9eOOeK3bXVNZ0OAX/inV9vBMWnosZklPbJA4H09sntVnUvGBsJNDurnzEtb21LzJCQSjEKQeRzjJ44/pXF7Wtze7rv+W3mdHJTtrocL4m0yztr7ytIs7z7Pbx7ZZ5FYrIw6sOOB79PQAdcrTtLutSuI44IJWRpFRpEjLBMnqa7bxLca8ujSX1lrg1DRLhdjMURXUNxhhtHrjjB9hVXwRNeReHvEbWRk+0JHE0QQbju+foPXit41ZKjzbvb/AIfQzcE6ljN8XaIljrc8emabcx2UKAM+12UnGSdxzxzjr2qyvhlT4EgvotNurnUrqY+WYldtkeepUcYO3rj+KtvSLzWbzQdcXxE93HAIkUPLBtKqSQ5AwMnFUfEWu6tDDb32izPbaEMW1s0ZA3lQedpGR0I5/u1CqVHamnqnvffQpxgry7nHXWl6hYor3ljdW6McBpoWQE+nIqOOyu5kDxWs0iHoyxkiut8Q3lzf+ANEuruZpppJ5dzt1OCwH8qsQaxfaR8MrGewnMEv25oywUHKkOccg9wK29tPlTsr3sZ+zjfytcg8FaMk41Uajp27bb7ovOi6Nz0zXKf2bf8A/Plc/wDfpv8ACvQ/BXiTV9V/tT7beGXybbfH8ijaeeeBWT4a8X69f+I7G1udQaSGWTDr5aDIwfQVkp1Yzm7LS3V9vQtxg4xRxLKyOyOpVlOCCMEGkrovEtubzx5d2ynDTXSxg4zjOBWsnhnSbb4h22iHzLq1aMmUSvhg2xmHK49F/Ouh14qKb7XMvZtuy72KusaJY3fhez13RYtkcS+VdwBtxRu7EnnqfyI6CuRq/qTvbXt9YwySJaLcviEOdvBIHHc4A5rXs/DltqnhCS/0+SSTUrVybiDj7nbA+gz78j0pRl7OPvPRv8+42ud6IxNP0y81W5+z2MBmmCltoIHA+v1rstB8Dzrp+pTavpTyTLGPssPm7Szc/wB04x061S+G/wDyM0v/AF6yfzWl0O01jWNLuNQk8TTWcEEgjdprh8c4xzn3ArKvUndxTslbv19C6cY2Ttfczv8AhCfEf/QLk/77X/Gqt/4Y1nTLRrq8sXhhUgFyynr06Gunk0fVP7OvLy18ZG7FpEZXWG5djgAns3GcGqwu7m8+F97JdXEs7jUAoaVyxA2pxk/WhV5trVPVLZg6cfMyvB+k2uta+tpeBzD5TOQrbckD1rsP+EQ0H/n2T/wYj/CsPwmNON1GdMbWf7UEJMnlJAUxxuxvPT6811Itb0nAi1Un2trD/GscRUl7TSVvwLpRXLqrle38F6BPOsZtgN3dL8MfyxzXmN3EtvezwoSVjkZAT1wDivWI4NQSQNEmrq46FbewB/nXnWurpKzOLI6j9rErC4+1iMDPfGzvmrws5OTTd/xJrRVlZWMaiiivQOYKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKAPVvAX/IsR/8AXV/50UeAv+RYj/66v/OivmsR/Gl6s9al8COD8W/8jXqP/XX+grFra8W/8jXqP/XX+grFr6Ch/Cj6I8yp8b9QooorUgKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAkghkubiOCFC8srhEUdyTgCug0HV/EOhanDY28dwSzgGxlUgN9Afu9+R+Nc7HI8UiyRuyOhDKynBBHQg10w+IPiMW3k/bE3f89fKXd/LH6VjWjKSskmvM0g0nduxualBZWXxYsRbeXGHZGlVSABI2fyJ+U/j71ia8RB8RJpphthS9jZ2I4A4PP4ZrnJLu4luzdvPIbgvv8AN3Hdu9c+taer+KdV1uzitb6ZHjjYP8sYUswBAJI9ifzrKNGUWuulmU6iafrc7Hxlqtnp2usLzw7b3YkRWjuZJCN4wPbHB/p61h3Gtzax4i8PK+nCxhgkiWBBnlC4wQTjjjA+hqjp3jTW9MtFtYrhJIUGI1mjDbB6A9aittXvNW8U6bd6jcebItxEu9gFAUOD2wO5qIUHBarZb3f5FSqKT0e5f8eWN0vjG4Y28mLkp5BC58z5VXA9TnjFanj2O5l1TS7a0SaW8trQO4hUsyc9ePcU3xH4y1bTfEeoW9hdxtbZXy8qrhCUGSp+tclFrepQaq2px3kgvXJLS5yTnsR0x7dOBRShUlGEnbRafcE5RTku7Ot8Ja9rmqa8mn3zveWcgZbiOaIMFGD1445GOfWua1qwjHim8sdMQyL9oKRInPP90Y9DkfhV648eeILi1MBu1j3DDSRRhXbj17fhisGzu57C7iuraTy5om3I2AcH8a0p0pRk52S02X5kymmktxtzaz2VzJbXMTxTRnDI4wRXp2l+IbnxHpI+xppra1DnzILmHh0zwUOc9Ovv6cVwtrqdne61NfeI0nu1lT5vIwrFhgDpgYwMVpv4us9OjePw7pEVizAr9qlPmS4PXGenPuelTXhKokuXVdeg6clG7vp+JveLvF1xpOsiwt7OxlWGFQxmgJ+b/Z5HGMfrXG674jvPEAthdRW8a24YIIEKjnGe59K04PHV3JCsGsWFnqkS9PPjAcH64x+lR6hq3ha6sJjBoElvfOmEKzHy1b169vpzU0qfsrJw17oc5897S+RoeCZrxNK1BglhFY22ZpLq6tmkO7A+UYI7DPc8+9dbLeXt7ZWDWkWmtdPpq3K280BbzDgZVPmG0c/qK8zsdddLa10y/wB82kRzF5IIztZgfcYzg8gE9f0u674jttR1HTzYJc2llZQiBNr4l29Dg5P8PH86mph5TqXt/X/BKhVUYWNLxFJIvgPThcxpDcXl9LdGFRgLkt0HYfMPzqt8Qv8AkK6b/wBg2L/0J6z/ABTr0Ot3lutnC8FjawiKCJgAR6nj8B16AUvizV7XWb2xmtC+2KyjhcOuCGBYkfqK0pU5JxbXf8SJyTTS8jf8I+Krq51fT9JW2s7a3cFXaCHa7kITkkcZJGeAKh1XWPC0mp3C3uj31xPHI0bSPduc4J6ZbpXOeHNSg0jxBaX9ykjwwliyxgFjlSBjJHc1qXPjJZLqZ08P6GyM7FWlssuQT1b5uvrUyoWq3itLdHYaqXhZs6HStV0//hHtXutChvLGSxiDKslwzplieQhJXse3euE1fWbzXL0Xd86tKECDauAAP/rkn8a308ZW8miapYy6VZ2sl1GqRmxgEYJ5+/zz7fjXI1pQpcspSa/Xp3JqTukkwrrrr/kldj/2Ej/6C9cjW/JrNnJ4Ih0YpP8Aaornzw20bO4xnOeh9K0qpvlt3Jg0r37GBWp4fvbLTtVS6vop5EjGU8h9rB+MHOR71l1e0ltNTUUbVY5pLQA7lhbDZxx9efcfWrmrxaZMd0dRHeeFNW1BUXRtVurudgBuuCzMfclq6fXLPQJNPWE2M9+NJxC8FtId8IYA5IyCw4HPPf0OORl8YWemwPB4a0tLLeMNcy/PKfp1x+ZrntP1e+0zUhf2tw63GSWYnO/PUN65ri+rzm+ZXVtrv+rG/tYx0et/I35NU8Hmzkgi0vUlU/Mqmc7d+MAkbverHgmS5h8PeJZLMyC5WOExmMZbPz9Krah4i0PWNPme90TytWKEJNbNtRmx95hkfkc/WqWheJ5NA02/gtrc/abrbtuBJjy8Zx8uDnqat05SpuKTvdbu/USklNO+nkdFYXmt3fhbxF/bDXRC26+V56Fe5zjj6Vz2ma/Bb+GNS0a9iklSfD2+0/cf1PPTIU8e/rViPxxqUml6hZagzXguovLRmYL5XXJwF56j8qzdCudFt5pv7asJbqNlHl+U5Uq35jg/0pxpNKXNHqnp8vQTmm1Z/eamvCSDwR4Ztyw2uJ5SB7sCP0Y0/TvGr6R4Zg060tUa4SUuzzKGQqc9vXkVn+JPECa49rFb2i2tnaIY4Iw2Tg46/gBx+prCq4UVKmlUXVv8WTKbUrxfkep+EPFF9rP9pfaIrVPIt96+VFtyeevNc/afEbUI7uJ7m0s2hDAuIotrY9jnrWf4V1+10JdR+0xTObmHy08sA4PPXJHtXOVnHCw55XjppYt1pcqs9TqNP1awvfH66testtamYy4kBbBA+XoOucH2q/od9/aXxXF2GVleaYIV6FRGwX9AK4itHQ9VbRNZt9RWISmEn5C2MgqQefxrSpQXK+Xe1iI1NVfvcj1n/kOah/18yf8AoRrX0ybWfCHk6sluptryLahdso+RkZAOcg/19auy+LtCnmeaXwjbtI7FmY3HUnqfuVl+JPEn9vmzjjsks7a0jKRQo27GcZ5wPQcYpLnmlCUdOt7f5j92N5KWpq/D2TzvFlxLsRN9vK21BhRkjgDsKi0HUNJPhDUdH1DUDZyXM6uriBpMAbT0H+7jrWb4W1uLQNVe8mheVWhaPahAOTj/AArEpujzTlfRafhcFO0V8zttPuvD+iaJrVvBrbXc17bGNF+yPHggNjnnruqva/8AJK77/sJD/wBBSuVt5FhuYpXjWVUcMY26MAeh+tdNrvimwvNHOk6TpQs7V5RK5LcsR7D6DnPaplSakrXd2m3p0GpprXTQX4fzG116W4a3uZYlgZGMEDS4JIxkKD6Gu20S2hj1iB0soY2G7DLocluR8p/5aFiF/r07159o+vadpdmI20q4kuGz5k0WoSQ7xk4yqjsDWtaeOrKyuUuItHui6Zx5mqyOvIx0IIPWscRSqTlJxT/D/MulOMUk3+Z0unrb2N9Fc/2eBsJ5g8PTRv0I4bJx+VeY6xu/tq+LRyRlp3bbIpVgCSRkHkV0f/CYad/0CL7/AMHE3+FYetajZalMktpp8lrJyZWkumnMh4xy3THP51ph6c4TvJPX0/Rk1ZRlGyf5mXRRRXac4UUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAereAv+RYj/AOur/wA6KPAX/IsR/wDXV/50V81iP40vVnrUvgRwfi3/AJGvUf8Arr/QVi1teLf+Rr1H/rr/AEFYtfQUP4UfRHmVPjfqFFFFakBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAereAv8AkWI/+ur/AM6KPAX/ACLEf/XV/wCdFfNYj+NL1Z61L4EcH4t/5GvUf+uv9BWLW14t/wCRr1H/AK6/0FYtfQUP4UfRHmVPjfqFFFFakBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAereAv+RYj/66v/OijwF/yLEf/XV/50V81iP40vVnrUvgRwfi3/ka9R/66/0FYtbXi3/ka9R/66/0FYtfQUP4UfRHmVPjfqFFFFakBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAereAv+RYj/wCur/zoo8Bf8ixH/wBdX/nRXzWI/jS9WetS+BHMeJPDes3niK+uLfT5XieTKsMc8Csr/hEte/6Bk36f417vRXTDMKkYqKS0MpYWDd7nhH/CJa9/0DJv0/xo/wCES17/AKBk36f417vRVf2lV7L+vmL6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/wAa93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8a93oo/tKr2X9fMPqkO7PCP8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/ABr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8I/wCES17/AKBk36f40f8ACJa9/wBAyb9P8a93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8a93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8AGvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/wAa93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8a93oo/tKr2X9fMPqkO7PCP8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/ABr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8Dh8NazceZ5NhI/luY3wR8rDqOvvUv/CJa9/0DJv0/xr1rwz97Wv8AsKTfyWt2qnmFSMrWRMcLBq9zwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6Kn+0qvZf18yvqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8AGvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/wAa93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8a93oo/tKr2X9fMPqkO7PCP8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/ABr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8I/wCES17/AKBk36f40f8ACJa9/wBAyb9P8a93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8a93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8AGvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/AIRLXv8AoGTfp/jR/wAIlr3/AEDJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/xr3eij+0qvZf18w+qQ7s8I/4RLXv+gZN+n+NH/CJa9/0DJv0/wAa93oo/tKr2X9fMPqkO7PCP+ES17/oGTfp/jR/wiWvf9Ayb9P8a93oo/tKr2X9fMPqkO7PCP8AhEte/wCgZN+n+NH/AAiWvf8AQMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/Gvd6KP7Sq9l/XzD6pDuzwj/hEte/6Bk36f40f8Ilr3/QMm/T/ABr3eij+0qvZf18w+qQ7s43wbp95YeH1gurd45RIxKketFdlRXBObnJyfU6Yx5UkgoooqRhRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQBheGfva1/2FJv5LW7WF4Z+9rX/YUm/ktbtaVvjZEPhCiiisywooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigDC8M/e1r/ALCk38lrdrC8M/e1r/sKTfyWt2tK3xsiHwhRRRWZYUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQBheGfva1/wBhSb+S1u1heGfva1/2FJv5LW7Wlb42RD4QooorMsKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiubvfHOh6feS2lxNKs0TbWAiJqD/AIWL4d/5+Jv+/LVqqFVq6i/uIdSC6o6uiuU/4WL4d/5+Jv8Avy1H/CxfDv8Az8Tf9+Wo9hV/lf3MPaQ7o6uiuU/4WL4d/wCfib/vy1H/AAsXw7/z8Tf9+Wo9hV/lf3MPaQ7o6uiuU/4WL4d/5+Jv+/LUf8LF8O/8/E3/AH5aj2FX+V/cw9pDujq6K5T/AIWL4d/5+Jv+/LUf8LF8O/8APxN/35aj2FX+V/cw9pDujq6K5T/hYvh3/n4m/wC/LUf8LF8O/wDPxN/35aj2FX+V/cw9pDujq6K5T/hYvh3/AJ+Jv+/LUf8ACxfDv/PxN/35aj2FX+V/cw9pDujq6K5T/hYvh3/n4m/78tR/wsXw7/z8Tf8AflqPYVf5X9zD2kO6OrorlP8AhYvh3/n4m/78tR/wsXw7/wA/E3/flqPYVf5X9zD2kO6OrorlP+Fi+Hf+fib/AL8tR/wsXw7/AM/E3/flqPYVf5X9zD2kO6OrorlP+Fi+Hf8An4m/78tR/wALF8O/8/E3/flqPYVf5X9zD2kO6OrorlP+Fi+Hf+fib/vy1H/CxfDv/PxN/wB+Wo9hV/lf3MPaQ7o6uiuU/wCFi+Hf+fib/vy1H/CxfDv/AD8Tf9+Wo9hV/lf3MPaQ7o6uiuU/4WL4d/5+Jv8Avy1H/CxfDv8Az8Tf9+Wo9hV/lf3MPaQ7o6uiuU/4WL4d/wCfib/vy1H/AAsXw7/z8Tf9+Wo9hV/lf3MPaQ7o6uiuU/4WL4d/5+Jv+/LUf8LF8O/8/E3/AH5aj2FX+V/cw9pDujq6K5T/AIWL4d/5+Jv+/LUf8LF8O/8APxN/35aj2FX+V/cw9pDujq6K5T/hYvh3/n4m/wC/LUf8LF8O/wDPxN/35aj2FX+V/cw9pDujq6K5T/hYvh3/AJ+Jv+/LUf8ACxfDv/PxN/35aj2FX+V/cw9pDujq6K5T/hYvh3/n4m/78tR/wsXw7/z8Tf8AflqPYVf5X9zD2kO6OrorlP8AhYvh3/n4m/78tR/wsXw7/wA/E3/flqPYVf5X9zD2kO6OrorlP+Fi+Hf+fib/AL8tR/wsXw7/AM/E3/flqPYVf5X9zD2kO6Orrn/GLanDoMl3pVy8M1ufMcKAdyd+o7dfwNVP+Fi+Hf8An4m/78tSN8Q/DbqVaaUqRggwHBFXClVjJS5G/kTKcGrc34nFeEdT17UteWzgv5EimlNxdEKvTjcenGeB+New15n4Z1zwt4dlvpEuZXaeU+WfJbKxjov15Ofwrof+Fi+Hf+fib/vy1b4uM6k/cg7ehnRlGMfelr6nV0Vyn/CxfDv/AD8Tf9+Wo/4WL4d/5+Jv+/LVy+wq/wAr+5m3tId0dXRXKf8ACxfDv/PxN/35aj/hYvh3/n4m/wC/LUewq/yv7mHtId0dXRXKf8LF8O/8/E3/AH5aj/hYvh3/AJ+Jv+/LUewq/wAr+5h7SHdHV0Vyn/CxfDv/AD8Tf9+Wo/4WL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8LF8O/wDPxN/35aj/AIWL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8ACxfDv/PxN/35aj/hYvh3/n4m/wC/LUewq/yv7mHtId0dXRXKf8LF8O/8/E3/AH5aj/hYvh3/AJ+Jv+/LUewq/wAr+5h7SHdHV0Vyn/CxfDv/AD8Tf9+Wo/4WL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8LF8O/wDPxN/35aj/AIWL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8ACxfDv/PxN/35aj/hYvh3/n4m/wC/LUewq/yv7mHtId0dXRXKf8LF8O/8/E3/AH5aj/hYvh3/AJ+Jv+/LUewq/wAr+5h7SHdHV0Vyn/CxfDv/AD8Tf9+Wo/4WL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8LF8O/wDPxN/35aj/AIWL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8ACxfDv/PxN/35aj/hYvh3/n4m/wC/LUewq/yv7mHtId0dXRXKf8LF8O/8/E3/AH5aj/hYvh3/AJ+Jv+/LUewq/wAr+5h7SHdHV0Vyn/CxfDv/AD8Tf9+Wo/4WL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8LF8O/wDPxN/35aj/AIWL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8ACxfDv/PxN/35aj/hYvh3/n4m/wC/LUewq/yv7mHtId0dXRXKf8LF8O/8/E3/AH5aj/hYvh3/AJ+Jv+/LUewq/wAr+5h7SHdHV0Vyn/CxfDv/AD8Tf9+Wo/4WL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8LF8O/wDPxN/35aj/AIWL4d/5+Jv+/LUewq/yv7mHtId0dXRXKf8ACxfDv/PxN/35aj/hYvh3/n4m/wC/LUewq/yv7mHtId0dXRWbpuu2OrWn2q0Z2iLFQWXHIorNpp2Zad9UeNeLf+Rr1H/rr/QVi1teLf8Aka9R/wCuv9BWLX0tD+FH0R5FT436hRRRWpAUUUUAFFFFABRRRQAUUUUAFFWEsLyRA6Wk7KwyGWMkGnf2bf8A/Plc/wDfpv8ACp5l3HZlWirX9m3/APz5XP8A36b/AAqCWGWB9k0TxtjO11IP6000wsxlFaGk6Lea1JcJZiMvBEZWVnAJA9B3NUooZZ5VihjeSRjhURSSfoBS5ldrsFmMorcs/B/iC9I8vS50GcZmHl49/mxxWRc201ndSW1xGY5omKuh6g0lOMnZMbi0rtEVFX9N0a+1dbg2UQk+zx+ZJlguBn3/ABP4GqFUpJuyFZ7hRRRTEFFWtN0+41XUYbG1UNNM2FycAdyT7AAmm39jPpt9NZ3KhZom2sAwI/MVPMr8vUdna5XooqxZWF1qVyLezgeaYgttQdh1NNtJXYJXK9FXdP0m+1S6ktrO3Mk0SF3QsFIAIB+8R69KSx0q+1MTmytZJxAnmSbB90f1PsOetJzit2FmU6KKuWml3t9a3V1bQeZDaKGmbcBsBzjgnnoelNtJXYJN7FOiprW2mvbqK2t4zJNKwRFHcmnXtlcadeSWl3GY54zh0JBxxnt9aLq9uoWdrleiiimIKK1NZ0G80M24u2iP2hPMTy2zx+VZdTGSkrobTTswooq/puj32rC4NlAZRbxmSTBGQMHoOpJxjAptqKuwSb0RQooopiCiir2paNqGjtEt/b+SZV3IN6tkfgTScknYdna5RooopiCita18Ma3e20dzb6bPJDIMo4AwRU3/AAh3iH/oFT/p/jWbq01o5L7yuSXYw6Kv6hompaUkb39nLAshIUsOCRVCrUlJXQmmtGFFXZdKvYdKg1OSHbZzsUjk3r8xGQeM57HtVKhNPYGmtwoq6dJvV0ldUMIFm0nlrIXXlvTGc9j2qlQmnsDTW4UVem0q6g0y0v2VTBd7/K2tlvkOGyKo0Jp7A01uFFSz209qyrcQSQsy7lEiFSR689qs6TpVxrWox2NqYxNICR5j7RwMn/IpOSSu9gSbdijRUjwSx3DW7IfOVyhQcndnGPzrUtfCmv3hIi0m6GO8ieWPzbFEpxirtgot7Ix6Kt6lpt3pN9JZ3sRjmTqM5BHYg9xSppd7JpUmppBmzjk8t5dw4bjjGc/xDtRzRsnfcLO9inRVrT9Ou9UvFtLKEzTsCQoIHAGTyeKgmhkt55IZV2yRsUdfQg4Ip3V7BZ2uMooq1eadeaeYheW0kJlQOm8Yypour2CxVoq9f6Pf6ZfJZXdsUuXAKRqwctk4GNpPWo9Q0670q7a1voGhmUAlWweD3BHBH0pKUXswaa3KtFFaGjaNc67f/Y7QxiXYX/eNgYFEpKKuwSbdkZ9FPmiaCeSF8bo2KnHTIOKZTEFFFX/7Hvho39reQfsZk8sSAg889uw4xk+oobS3Gk3sUKKKKYgoqxZWVxqN5HaWkfmTynCJkDPGep4pLyznsLuS1uY/LmjOHXIOD+FLmV7dR2drkFFFKqs7qiKWZjgADJJpiEorc/4Q7xD/ANAqf9P8aP8AhDvEP/QKn/T/ABrP21P+ZfeXyT7GHRVi9sbrTrlra8geGYAEo45waSzs7jULuO1tYzJPISEQEDPGe/0q+ZWv0Js72IKK6D/hCfEf/QLk/wC+1/xpD4J8RgEnS5MD/bX/ABrP21L+ZfeV7OfYwKKKuJpWoyxrJHYXTowyrLCxBHqOK0bS3JSbKdFXf7G1T/oG3n/fhv8ACj+xtU/6Bt5/34b/AApc8e4crKVFOdHjdkdSrqSGVhggjsabVCPVvAX/ACLEf/XV/wCdFHgL/kWI/wDrq/8AOivmsR/Gl6s9al8COD8W/wDI16j/ANdf6CsWtrxb/wAjXqP/AF1/oKxa+gofwo+iPMqfG/U6nRtA0waA+u65POlr5vlwwwY3SkfX6EduhNUdah8PtaR3Wiz3KyF9slrcKMqMcEEcY49T1HStDw74ms7fS5NE1u2afTZG3KyfejJPP4d+ORz1zS6z4aj0uyg1/RL/AO1WHmKVZgN8ZzxnsefYY44rBSkqnvtrXTs/I0snD3fn3MvUPD8un6Dp+qvOjpeZxGAcp6c9+P8AJrIVSzBVBLE4AA5Nem654m1ez8HaLfwXey5uf9a/lId3HoRgfhVfwppdzbeD5NS0xbf+1LqQpHNMQPKQHBxnjOQfzHpSjiJRp80+9v60G6ScrR7HnckbxSFJEZHHVWGCKbXpup2d9deENQPiSSzkurYeZa3KOu/ryvAHXAHvn2qnDqj6N8NdLvIIYnuvPkjhkkXd5RLPlgD3wMfj+FUsS2tFd3tvoJ0rPV9LnASRvE5SRGRx1Vhgitfw1oi61qLLcSGGxgjMtzMDjYo9yMZ+vofSunFxN4u8CX0+oqst/YyZhmWMBiDg7cDHXkfl3FQ6tp99ofhm30OytbiW5uv31/LDExHtHnH+ce9Drtrk2le3/BBU0ve3Ri+L9DtdA1eO0tJJnieBZcykE5JI7Aelc/XbfEuCX+3Lefyn8n7KieZtO3dljjPr7Vh3uvJd+G7LSRYRRPbPuM6nl+vbHvzz2q6E5Spxe99yakUptbG94UvdYv8ATLtRr72FnpsSt8tssny/Nn0PGPeuj8N6qLu7uk/4SqTUNtq7+W1iYvLwR8+e+PT3rA+G84T+1beO4hiu540FuJOdzAP2746muv0q8uy941xrelXEcMLbvIQL5TerHPAGDXBibKclb8F/l+p00fhTv/X3mFpt3d6vctb2HjiaWZUL7Tpu3ge5PvXCNqupajrUF9KTd3iMpUGPO7acgYXGRXp9hrIinZr/AMSaLNDsICwlUbPbnNeV6PPNb6xaSW9ylrL5oUTuAVjzwWOeMAGt8PvJ2/r7kZVemv8AX3s9Cl1vW4rbSp7C0sHmvgUmX7IyNE+7GD82dvv7Z7inXd5dN43sdOW1j+xW90refHbFPn2EY3ZII59uR7c6OhT373VyJ/FVjqIFuxWOEJmM5HznHYf1rNtrnUX1fS45vFFlqkb3QDQQBNwwrHdxzgY/WuVWu9Fon3/y6G2un9fqYutaX4vn1m/aBdRa2a4kMQWY7dhY4wM9MYrk7y0vo9Sa1u45Te7gpRss5JAwPc8iu/0/VL+S+8Zhr65ZbeK4MAMrYiwXxt5+XHHSuL02yfxBqErXmrQwMF3yT3khyw4HU9T0GCRXdQnJX5rWVtjnqRTtbqd7ocel+GBH4fvWik1DUU3Tl1BjUkYSM+o68e59RTLJYze3a6n4V0uysrIE3FyYsjAGQE4+Yn/POAYLXwlY+QNSC3XiOZycFJFjiZge5ZsnpjqfpV2b+3tYVrXXvC4msd+6MQXCq8WM453fMfy7/SuOTi22n662+5XX4m6Tta3p1+8861u/t9S1SWezs4rS2+7HFGgX5R3OO5qnb2813cJb28TyzOcKiDJNdfrvhvw9p8Fxt1G6tb1IxItpcIGJz0XK8e2QTjrWbo3i6fQ9JmtbOytlupD8t3t+dR7+uO2ePY16Eal6f7pX9TmcbS99nQ6dp8nhKz8iILN4n1FNsMKsv+jrgnOTxnj6EgAdCaZf6bH4ximaNY7bxNZjZdW24Yn28ZBzjPb2PB7GsDwvcz3njewuLiVpZpJ8u7HJJwaq6zcz2fi7Uri2leKaO9mKuhwQd5rH2Uvab+9a9/09DTnXLtp/X4mfd2dzYXDW93BJDKvVXXB/+uPeul0JfD9itrfP4hura/XDOkcDYBz904HI9fWqureMtQ1rRo9PvI4SyuC06rhnA7H055yMVt+HdDFto8Go/wDCO3OoX7kvF5kqiEL2bGfQ9CDyO3FXVk/Z/vNH5NfqTCK5vd1/ryOvsNO0+51KTxFbzsou7ZowvkmMMOMvg8np1+lYk1no+meEIbMald2tjdkyNcx2rbplPQMQOB7cZH45k0G18TXPiO4vtZtmiRrR4o+RsjyVOAoJ9KydAs/F2jD7LNpL3emPkS2sjIRg9cZPH06GuFRs37+1uq8+vkdDf93e5yWqQaTb6hGmnXk11aFQXkaPawOTkAEDtiu7u7bw34c0688O3Op3kZuWWWRhHlscYAIGMcVxfibTpbDV5GOnPYW8zFoIXcMQv4E/l2ziuz8ZQeGJPEDNqt5fxXXlr8sKArjt2NdVV83Iru2+nlYxgrc2iGad4Zs/D/iqW+M0rafp1r57SSAZLsCAOBzxz9cVVh0/wh4j10omoas93dyM/O0DPLHqvTirPibVhB4wh029cvo83lPPCcAE4wGJ64GAcZ7VU0rRv7D+KlvaAfuCZJICf7hjfH5cj8KyTk4ucpO/LdfIt2vypaXOa03SIdQ8VLpRkeOFpnjD8FgBnHtnitm50PwhaXU1tNr92ssLmN1+zE4YHB7VT8NMtx4/tZIzlHundT0yPmNamr+GNLuNav5pPE9jC8lxI7RMOUJYkg89R0repO1RRcmtOn/DMzjH3bpX1NnxhZ6BP/Zv9p6pNbbbfEWyEtvXjk8HFczq3hjS4PDJ1rS9RmuYxKI/3ke0HsfQ10/i/RLHUf7N+0a3a2XlW+1fNH+sHHI5FUtVsoNP+GLW9vfRXsYugRNEPlOT0rnoz5YwSk99un5GtSN3K66Hn8VrcTxSyw28skcQzI6ISEHuR06H8q29FbU/DUtj4gMOLGZ/LLK6t5i5+ZcA5B4PXuK19OZtM+FmoXBZVe+n8uPJ6rwp/HAeodV/5JboX/XzJ/6FJXVKq5vlto3b8DFQ5detrmFqTRa14hlbSLKVRcyZSD7zEnk8Dp3OO30rMdHjdkdSrqSGVhggjsan0+/uNL1CG9tWCzQtuUkce4PsRxXS+NV0y++x65YTRiS9TM9vv+ZW6bsduhB9x7mteZwkoW0ItzJy6l5rPwro2i6Pd6lp91PNeQiQmKTjIxnILD1pdV8S+EdavBdX2m6lJKECAhgoAHbAb3NJqwDWngdWAIKqCD35jqfxR4sudG8RXWn22n6a0MWzaZLfLcopPQjua4opya3b16262N20k+i06eRSspfA99f29pHpOoq88qxKWl4BYgDPz+9YeuaSYvE17p+mWs0iQvhY4wzsBgc+tdR4mjjj+JGhiONIwTbEhFwM+aa5/wAT3dzZeNNTltLiaCTzCN8TlTjA4yK1otuScW9VfV36kVEkrPoztdPvYYdA021mt3E8EW2RJ7K4O0+21cVN9utv+faH/wAAbr/4ioI3WDQtKuZr9vOuYA7td61NBk4H3QM56+2KmuHgh0+zuft0QM+/JfxBOqHacfI38XvwMVxNK/XVs6E3Yx/GDNqujWNrYWlxLLFKzOkNnMqgEf7SisfwX5yzanbSaVNe28sHlTpEg3oTnHUjH8X5D0ra8UzzWfhuyv7DUbmOSacozQanLMhADdGJ9vSuc0Cze5t57keJotLkaTDq85RpOM7uCM9T+tddP+A10/4PkYy/iJnUroV5/wAIhoVpd6bczfZ7tpJ7eMDdty57nGDkfnVOySPWfGutnUdOjj8uydkgkjXKYKBScfxYPX3rWvbSUeEtOgbxSkMnmswvWmYecMnjOcnqO/asrw3aC08Tamv9qRakz6VI7XEb7+dyjBOTzxWMZXjKV9de/ctrVL0/IzLr/kldj/2Ej/6C9dB4anmtfC2gzwW09wVvZd8cABZl2yepAxnB5Nc/df8AJK7H/sJH/wBBetfSrK91DwRodtYzTQSPfPvliYqUT59xyP8AOcVpVs4O/wDM/wBSYX5tOyNO3sbmwtdHuF0i5ke0W9aKAAAxu75jDc8ArnpmsW41fXrLwros1g0ouJ2uTcstursSJOM5U4xk+ldLqV5Z6t/Ztrb3txCLuO5itbiKdlHmRlVBOD8wODjJ/nXPS6hquhaP4as47popJZ51uNpDb/3wwc/ifzrKneVrpXvs/wDt79S5WWz0/wCGOU1e91zVWhbVBcSFTsiLw7eT2GAM9Oldp/b+sxaBDqkNlZfbROYpLc2bK6jHBHzZPuR6+xrF8c61qA8UT2ouT5FpMksCFR8jbAc9PUnrXX6JPqDa1As/i+w1CM7s20Ij3P8AKemPTr+Fa1n+7jJxVt7f0iIL32k2ZfiO91GO4sdPazgbzRDLdyQ2xAjfeCcPn1HcdD78V/Fun+KbnxJdvp4vzZnZ5YimIT7i5wM+uas6jc6nvdf+ErsbqNrhENkgj3kGQDb68Z/Srtrf3b/Fa8szdzm1WIEQGQ7AfLX+Hp1zWUW4LmSWiffy9C2lLR31a/U8/XSrxvEdjp+tC5ie4dEyzZcKzYBGe2a7K6t/Da2kPhJ9TvEeK652x8u7HgE4xj5q4zR7y6vvFeky3dzNcSC7hUPK5c43jjJrsb2DwwfG++a8v11H7Wh8tUGzfkYHTp0710V78yTvor6dzKnaza/EXRNKs/Ceo6zqlzNKLO2ItYZCMsxbaWIGOSDgfnWdY6F4R1iW4SyvtVknjheY+YVGQPfb6mrVxfRX/jufQ9XJl00zOsMeQoSVhw2RgnqQOerVS8L2TaJ4k161ujj7Lp8258dVBUhsc9Rz+NZ+9ZybfNZP5f1+Zel0ktLnN6NbaTczSjVr+S0jVfkMcRcsc+wOK9BtLfRvEui/2LHq1xdi2PmC5lt23xL6bmGP/rD244nwpo8erak/2izubq2hQs627KDn+EHJHBwRxXS6taeJrmx/svS9BbTtNHBjSRS8g/2mzzn/ACTWmItKfKpWa81ZfqRSuo3t+Z1NxY2H/CUSamZJJLu1s1/diAuEBztcY6tweB71514gh0KWO6u4dcu7zUmcEpNAV3HIBycDGB29sV0/ijS/En/CTw6losMoZbZULxsowcnKkE89u2KzPFdlqOo6XHfXPh02l7Flru6R12soGBwDk/j0x3rHD2i4vm3819xpV1TVvzMvQPD2m3+g3mq6lfTWsNtKEYxpu4OMcdeprpfBtl4cg8QK+l6tPc3PlsPLeAqMdznArO8LWUV94D1qC4u47SKS4jBmk+6uCp5+vA/GtHwboNhp2vrPba9aXsnlsvlRDnB79TVV53VROT9Om3oTTjZxaRkT6R4SuNTlRtduhPJMQUFucBi3TOPWsTxRokfh/WmsYpmlQIrhmGDzXQf8Izpf9t/aP+EmsfM+07/Kxzndnb169qp/Ej/kbX/64J/WtaU/3iipNq3X/hkROPuNtdTlzZ3IgjnaCRYJW2pKykIx9mPHY/lXU6bfXHhA3uj69ZO9rdwbjbo6k7jwDkHjIB9+BVnxuzWGgeH9HLLvjgEkqg8hsAD8M76h+Jn/ACNS/wDXsn82qlU9taLWjv8Ag9BcvJdrdW/E5WKzuLpLiW2t5XigXfIVG7Yue5/z3p2mwpcapaQSjMckyIwzjILAGtbwhr39iawBMQbG5xFcqwyMdj+GfyzUl7p9ppnjm3t7G4jmthdRlNj7tnzDKk+oP9K1dRqTi100IUVZM6SS+8J+EPEUiR6Zem8tuN6sGX5l7Zb0NZU+q+CLm4knm0rU2llcu7GQcknJP363rKKOb4t6mssauvkZwy5H3Y6oeGvFtzrHiG0sLnT9NWGYsGMdvg8KTxk+1cS25tW0k27nQ97ab9hmk2fg7XJLqCz0y9jlht3nzLKccYHZj6iuR0myu5ryC4htLiWGKZS7xRM4XBBPQeldZ4ZUL408SKoAUW90AAOn7wVy+h317DqNrawXt1BBNcIJEhmZA2SAehHauiHMuZJ9FuZSs7XPTrvU7Sa6kkSBWVjwZLG53HjvhKh+3W3/AD7Q/wDgDdf/ABFJqMkFlfy2wvogEI4uPEE8b9AeV5x+dGpPBY6hLbC+iATHFx4gnjfkA8rzjr6156itEk/wOpt+RyPjhJtR1t7+1tLprRYVVpWtpEUEdfvAVR8E/wDI46b/AL7f+gmtLx5NdafrH2C3v737LJbqzxPdSSKxJOfvE56VB4aTQFNnK+oalBq7OUAt1XAYkqACR3BH513xf+z27rT/AIJzNfvTpr6azGoXIbTPFTsJWy0LPsPJ5Xnp6Vf8OyW738oistegbyG+bUCxj6jgZJ5/+vUMFxBPrY0f/hINdjviSPLkVBjC7uu30rPm8QaTFLJDJ4m1zcjFWGxOo4P8NcbjKS5Un+JvdJ3/AMjzWur0LV/F2o+Vp+lXUhjhVUGEQLGvQZOP/r1n6qvh+2+zyaNPd3EgfMi3SDZgdugzmussr/W/EukzR6H/AGdothAdjqGKkkjn5gvA57YNd1ad4J8v39Pkc1ONpWv9xt6jeahoXg2+uP7VF9qMTqjylRiJmKjAA7gMDz6/hWDqP/CeadbpdR35vIGQMXtlR9vTtt5+oz0zVu18NyweB77RG1LTTcz3IlVhOdgA2dTjOflPas+y0nxL4Vsbi/tNW0+a1gXfJAsrSKfw2jB+hFcdNQV7Wbv1W+33HRLmdt7W7nCXE8t1cy3Ezb5ZXLu2MZYnJNR1pa7q39t6rJffZo7feANkfsOpPc1m16sb8qurHE9z1bwF/wAixH/11f8AnRR4C/5FiP8A66v/ADor5zEfxperPVpfAjg/Fv8AyNeo/wDXX+grFra8W/8AI16j/wBdf6CsWvoKH8KPojzKnxv1Ok0vTvDepaZGtzqsmnaimRIZI90bjJII/AgdR06Vb1fVNJ07wt/wj2j3DXhlm8y5uGQqpIx90H3VfUYHfOa5Cik6N5XbbW9h+0srJHcQ3uja94RsNP1DUxp8+nuS2Yi3mJz933xj8QeOlU9G1XTL7w3J4d1iY2qrJ5ttdBNwVj2I/E8+hPSuToqfq6s1frf0Y/avex0GpaToWn2Mxh137defL5UcUJVevO48jpnjiultodLuPhnpsGq3D2ySXDiKdVLCOTc+CR3GM/8A1q86rWuNdkuPDVpophQR28plEgJyxO7jH/AqVSjKSirve99Bxmk27dDeutV03w74bl0fR783t3cSiSa5VNqoBggDPXoB379OlZv/AAnXiX/oJt/36T/4mudoqo0IJe8rvzJdWXTQ7Hxv4p/tedLWwvTLp5iQunl7f3gJz1APp7VjX0Whr4espLO4lfVC3+kxsDtA59sccdD3rHopwoxhFRj0/H1FKo5NtnS6L4X8StPa39hA1v0kjuGkVQARnPqQQfQ9a7K+1HQ5ba60ybUrO01W+iVLu5toyYiw6gnp6jn1rzF9QvZLZbaS8uHt1ACxNKxUAdOM4qtWc6Eqkrze3b9S41VFWijrJfh9q7RmWwms7+I8o0Ew+YdjzgfrWSNEex122sNbb7CjkGR8htqn6ZxnH4dazYZ5reTzIJXif+8jEH9KJp5biUyzyvLI2Mu7FiccdTWkY1NpS09NSG4dEep6ddvqF7d6jb2wg0G202W2tpDtGcFck9x93v2HvVTQdQtNXuIvENxZtBJpVoy3Ew4R2wQoUDqdu76ZA54NedJe3UVq9rHczJbuctEshCMfUjoegpq3VwltJbJPKsEhBeIOQrEdMjoegrD6po1f/huv3mvt9jrfCkz3Nn4rnkxvlsJXbHqQxNcbVi1vrqySdLad41njMUoU8Op6g1XrohDllJ9zKUrpI9BtrOe9+HWlpFqdrp5W4kPmXE5iDDLcAjqc9qzP7Cv/APocNI/8Gbf4Vy7Xdy9qlq9xK1vGdyRFyUU+oHQdTUNZxoyV9d2+hTqJ20Ou+JH/ACNr/wDXBP61yNS3FzcXcvm3M8s0mMb5HLHH1NRVrShyQUexM5c0mzc8Hf8AI3aZ/wBdv6Gq3iNSvifVQwIP2yU8j/bNUIJ5raZZoJXilQ5V42KsPoRRNNLcTNNPK8srnLO7FmJ9yaOR+05/Kwc3u8pHXUaTonizULeFrR7uG0Zfkke4KIF9cZzj6CuXq1/aV99jWzF5OLVQQIRIQnJJPHTvRUjJr3bfMUWk9TtrnXx4SsTZ2eqNqeqySK00zuXjjVT9wZPfocc8npxSmFvFSteaDrU9reP80unTXLDae+w/3fwxz26V59Tkdo3V0YqynKsDgg+tY/VktU/e7/8AANPat6PY0tWstYg1CKz1TzjdMB5ayy7zgnAwckdRXceMrTw1Nr7Pquo3cFz5ajZFHkY7c4NecXFzcXcvm3M8s0mAN8jljge5qKqlRcnFt2t2EqiSate/c674jqF8UKQchraMj9R/StjwtomsafrDa1rg2QwW7FZp5lc9OMHJwMZ/OvOanlvbu4hjhmuppIowAiPISqgDAwD04pOhL2app+Q1UXO5tEUcjxOHjdkcdGU4IpGZndndizMckk5JNJRXQYncfEYEHRQev2T/AArifNk8ryt7eXu3bM8Z9cetSXF5dXZU3NzNOUGF81y20e2ahrOjT9nBRZc5c0rmhPrV7c6NBpUrobSB98ahACp57jr1J59a7SDRLvxD8N9Jt7AxNLDPIzh3xj5n/wARXndFTUpcyXLpZ3HGdr31PQNF8ASWdxPceIVt/sCQsWxMcg8HOR0xg1wMhRpXaNNiFiVUnO0dhnvTaKqnCabcpXFKUWrJWOy8USvBoHhKaJtskduWU46EbCKz/wDhOvEv/QTb/v0n/wATWA80siIkkjsqDCBmJCj29KZUxoRUbSSe/wCY3Ule60O00Cz1vxVrVnrU88U62dzGsjMQrBVYN0A9zWbr9rNq/ji+t7BRPLLMwQKw5IHPJ47GsGOeaKN445ZESTh1ViA319ep/Oo6I0mpuSfSy0BzTjY9ctbTVLO0htoIfEKRxIFVRNaYGOw4rRuF1P8As2y2JrXm/P5gjlthJ143kjafbb2614lRXO8E27tr7v8AgmqxFun4/wDAPS/EWj6xq+miIW2sSyxuHQXU1tsHqflwc4rjtA8M3viCRjCUitoz++nkYYQd+O5x/wDrFYtOV3VWVWYBxtYA/eGQcH8QD+FbwpThBxTX3f8ABM5TjKV2jovGGrWl7dWthpxBsNPi8mJv7x4yc9xwPyrI0vUb3TrlmsJlhknQwszBcFWIyCW4HQc1SorSNOMYcnQhzblzHc65pt3pHw3srO+iEVwNQ3FN4bgq/oSKXSdX0c+EbTTrrWLiwuIpXdjBGxODuGMgY6NXF3F3c3ZBubiWYr0MjlsfnUNZLD3jaT1vfT+mae1s7pdLHoDW/hrWbLTtFt9Wmd7cutsEtX3MXOTknjrzntVTxPPDFr+i6PbvvTTFjiZvVyRn+Qri1ZkdXRirKcgg4INOMshm84yP5u7dv3fNnrnPrRHD2lfmutfvYOrdbHReNUR/HV6ksnlRtJGGfbnaNi5OO9djo0gutb0q20WHfomlGRZbr5fnkMZGfXq3bg5PoK8smmluJWlmkeSRuruxJP4mpIL26tUkS3uZoVkGJFjkKh8euOvU/nSnh3Kmo32Vv0uEatpNnoWjXVr4l1KzgltP3mlXDy/akOI/KByucdSWx+ROeoqp4UvxqnxKur0crKJSnGPl4C/oBXDQ3VxbrIsE8sSyrskCOVDr6HHUU+yvrrTrlbmzneGYAgOh5waTw2kknurIaraq5c8NRvL4o0pUGSLuNvwDAn9BXb3tp4Zbxt5s2pXa6j9rQ+SI/k35GBnHTp3rzSitKlFzlzXtpYiNTlVrHReKopj46vI7cnz2uF8shtp3EDHPbnvW/puj6lo2jeIdR1sCKW4s2hR5pQ7uWBHXJ68V59U9ze3d6wa6upp2XoZZC2PzpSoycVBPRfoNTSblYuaNaazeTSR6OtyXABk8lyvHbJyBXX2dlceHSmpeJtZmBT549PS5Z3kbtnnGM/h6muEtb26sZGktLiWCRl2F4mKnGQcZH0FRSSyTSNJK7O7clmOSfxp1KUpu17L8RRmorzO6g8Tx+Jlez1S/l0u5Ds9tdwyFUwT9yQDAIHGDx07c5zdd0PxNpdrLLdXc1zYYGZkuSyMCcDIJz6dq5Wp/tl0bT7J9pm+zA7vJ3nZn1x0pKhyS9zbt/kN1OZe9uRiaVYWhWVxGxyyBjg/UV1fw2B/4S1faB/6VyNTW91cWkvm208sEmMbonKnHpkVpVhzwcV1JhLlkmSal/wAhW8/67v8A+hGoWmkebzpHMkmQd0nzZx6560xmZ2LMSzE5JJySaSrSsiWy/q2sXmt3SXN86vMsYj3BQuQMnoOO5r0Dxj4S1PxBrMd/p5geAwKoLSY6En+teYUVjOi7xcHa3kXGas1LW52954Ug0Hwfd3GsRxHUDKFtdkp74H49zj27Vyujf8hzT/8Ar5j/APQhVKlVirBlJDA5BB5FVCnJRak7tilJNqysdb4p1S90jx7qVzYzmGYhV3BQeCi+oPpVP/hOvEv/AEE2/wC/Sf8AxNc/JI8sheR2dz1Zjkmm0o0IcqUknZdhupK7adj0XwrpF/aDUPEN/LD9nu7CRvMDjJZiG5Hboa5XwzpF1qeprLBb3EsVsyvKbd0V167SN5x1FZDTzNCsLSyGJOVQsdq/QfifzqOpVKS5nfV+Q3NaabHsO3Wc8J4i/wC/1p/hVzVF1M6lL9nTWjFxtNtLbCPoOgcbvz75rxKisPqWt7r7v+CafWNNvx/4B2XjLR9RfbqkttqHlxoEllvJISRzhQPL7c+lHgzSktN3ibUj5VhaAtETj96/I4+h/XHvXG1J58v2fyPNfyd2/wAvcdu7GM46Zx3rf2UvZ+zv+HQz51z81j0zRF0bWfFcHiGz1Aw3GXM9lcEbslCuVPpznv8Ah0qWa0+IH2mVre9tlhLnZ8sX3c8fw+leVVNFd3NuhSG4ljUnJCOQM/hWTwjvdNPpqrlqsrar7mbnifQtX0+QahrE0DT3Uh4RhuYgcnAAGOn5it7wZbx3XgvXIJbhLeORgrSv0QY6muEmuZ7jb588ku3pvctj86dFe3UNvJbxXM0cEv8ArI0kIV/qOhrWdKUqai3qRGaU+ax03/CJaR/0Nun/AJf/AGVb1nplrpngTXo7XU4L9XXcXhHCnHQ8mvNKmju7mGCSCK4lSGT78auQrfUd6mdCckk5fghxqRXQhooorqMT1bwF/wAixH/11f8AnRR4C/5FiP8A66v/ADor5rEfxperPWpfAjg/Fv8AyNeo/wDXX+grFra8W/8AI16j/wBdf6CsWvoKH8KPojzKnxv1CnrDK8TyrG7Rx43uFJC56ZPbNMr0Pw/pEp8A3S2d5aR3OpPhmkfbtjBwVPGc/e/OitVVNJvuEIc7scPHpl5LpsuopAzWkLhJJQRhScfj3H50y4sbq1ht5ri3kjjuE3xMy4Dj1H+e49RXoz+Fru38DppEV3YJPcXHnTO8xCMo6bTjnovaoF0TVm8NSaLdXWjXMYYNbyyXR3QeuPl59vTJ6jiudYtPXTf8DV0H+B5xUstvPCkbywyRpKN0bOpAceo9astpNx/bX9lRNHNcGURBom3KT9fQf0r0HxPpNndwafpI8QWNpFp8QjMU0o3M2AMkZ64H6mtqleMJRXciNNyTfY4S90G7sNJs9SmaI293/q9rEt0zyMVl16lrWhW1z4S0ezfWrKCODO24dvkl4/hrjtU8O2Wn6dJcxeILG7lUgCGFgWbJ+tRRxMZrXe/ZjqUnHY56itTSdbk0iO4jSzsrpJ9u5bqHzANucYGff9BW/FqMOq+EteuJNH0qGa2WERvBahCN7kE59cDitJ1JRe2mn4kRin1OXj0y8l0ybUY4C1pC4SSXIwpOMDHXuPzqpXo9xo9laeErLQr3WYNNuWb7Vcqy5LE5wDg9uB/wEVg6/wCErTQrR3bW4pbkIrpbeSVZ1Jxkcn3P4VFPExk7Pvpoy5UmldHLVpaNol5rt09vZeX5iIZCHfHA9O5610XhJri+spoE0rQzBZxM73l9a7snJIDNn9ewFaza6ul+EbTXrLStIjvZbpoC0VttXb8/IwQ3O0dT+FTUryTcYrXYcKaau3oea0V6pPtu7jSI7XSfD4udQsvtUjXVr95yAcLg55J9zjJ7VwfiG8uJ7sWt1pNjp09szK6WsHllicfe5OenB9/erpV/aO1vxJnT5Ve5j0UUV0GQUUVd0q/j06+W4lsoLxApBinXKnPf60m2loNbj9I0a71u5kt7MIXjiMrb2x8oIB/mKh07T59U1CGytgpmlJC7jgdM/wBK7vw34u0o38/naVpumKLdsSouC5yPk6d+v4VcTUbPRJNJvr/R9JsUuhv3RRnzoRg84C/QfjXHLEVE2uX0N1Sg0nc8zubeS0u5raXHmQu0bYORkHBqKuzvfGOmtf3DJ4b0ydDKxWZ4+ZBn7xyM5PWsXQ7OLWvFEEL+VBDLMZChHyhR8xUfgCK3jUlyuU1axm4K9ou5mizuDeLaGF1uGYII3G05PQc9OtWDompjUZdPFlK13Epd4lGSBjOePYj8xXpE/h671Xx5Fqs13ZSWkEiskSTZdVTkcY/vcnnvWfPomtr4qn1uz1XS45HlLKGuDgp0CsMegGa51i0+234mnsGvvPOKlgt57pylvDJK4G4rGpYgevFdR4w0aZbifWGbTIY5XA8i2uC7Fj1PIGSeScYrR+HtkLO2v9duJ4raMIbeGSZtqbjgkn8Qo/OtZYhKl7Rf0yFSfPys5HRtHuddv/sdoYxKULjzGIGB+FUpomgnkhfG6Nipx0yDivSvCGhW9j4j+1pr9lfzMj7kifc5z1PWsW78J2Et/PI/ijS4y8rMVLjK5PTrUrFR9o09rdmU6L5U+pxlFWtStIrHUJraG6juo4yAs0f3X4zxW1J4ylmkMkuh6G7nqzWWSfxzW7lKycVcySXVnPQxSXE8cMS7pJGCIvqScAVLfWNzpt5JaXcRinjxvQkHGRkdPY16Euk2v/CyLy4MFta6bpMaSvsQKoJQFcgd8knP+zVO48K2Ou3moaknieCXBaaYiEkRqc993QDj8KwWKjdX2t+Zr7F203OBorVPk6L4gxaNa6rHEcIZYd0cu5f7uexPr1Fd5DH5GtaRY6ho/h0NfJukhis8SxcE88kY4Az9eOM1dSvyWdr9SY0+bqcBqWiXmlWtlcXPl+XeIZIij7uOOvbuKza9U0/WRqeuarotzpemNZaak726fZ84KOFHBOOhPQCse8vLy00O01hfD3hue0nX5mis8+Uc8Buf8nj0znDET+GS19e+xUqUd09Dirayu71mW1tpp2UZYRRlsfXFWf7C1j/oFX3/AIDv/hXRfDYldevmBwRYSEH/AIElX7K+1xbHQdTl12aVL+9WFrcoBhQ5B+bPPT071VSvKM3FW/r/AIYUaacU2cFNBNbTNDPE8Uq/eSRSpH1Bq7Lot1FoMOssY/ss0vkqA3zbue3p8pruYLO2vviVr0d3bxTotqWVZVDAECMA896x77/klGm/9f5/lJSWIb5UutvxQ/ZJX+f4HMabpt3q16tpZReZMylgucdBnqeBVeWN4ZnikGHRirDOcEda9C8H3On22kXGpQaIUuolS0WQXTE3Mj4GADwnO3ntmsi5OlQ6hJbP4NmWW3GJYo76R+TjGSAccZpqu3Nx5dF6f5+gnTXKnf8AP/I599IvY9Hj1Vo1FnJIY1feuSw9s57H8qpxxvLIscaM7uQqqoyST0AFelaxJYWvh3R7Q+Gp7iCRGuBbpcyAwk88kDJJ3HriuO1F0iu7O60zRrnTHRgRvdpN0gORgsPpxTpV5TWq79v8wnTUepU/sLWP+gVff+A7/wCFRz6TqVtC01xp93FEv3nkhZVHbqRXoI17VtC8LyXetXTyalecWls+FMa/3mAAI65x7AcZNZniUa8fCttdyapPd2VyALmN4QhjfOQCAOBkAdeo96iNebkk7Wvbrr6DdOKWlzkdL0241fUYrG1CmaXO3ccDgEnn6Cq88L21xLBJjfG5RsHIyDg16R9jtbH4t6fBZ28cEQhY7Y1wMmN+a891L/kK3n/Xd/8A0I1pSrOpLysn+ZM4cq+ZVop0kbxOUkRkcdVYYIrW1Dw/Lp+g6fqrzo6XmcRgHKenPfj/ACa2c4ppPqZqLZj0UqqWYKoJYnAAHJpZI3ikKSIyOOqsMEUxDaKKdJG8TlJEZHHVWGCKYDaK6qytvB9rbWx1G8vbq5lRTKtuuEhLAHnucdDgnvxVbxb4bTw9eQfZ52mtLlN8TP8AeGMZB9eo/OsVWi5cuqNHTaVznqKKK2MwooqxY3K2d7DcPbxXCxtkxSjKuPQ0nsMk0zTbjV9RisbUKZpc7dxwOASefoKj+xTf2l9gwvn+d5PXjdnHX6122geLtLOt2wl0TTNPT5s3KLgp8p9u/T8a0ri/tbW1i1ubRtHS2lvCqSlD5rLu4kHy8ngt+HvXJLEVIys4/wDDmypRcb3POtT0640nUJbG6CiaIgNtORyAev41Uru9a8Z6VJq9w8Gh6ffRkjbcyxYZ+B1yM8dPwrm4n0zVddkmvmXS7NwW228ZYKQMAAAdz7VrTqTcbzjbT+tCJRinaLKU+nXltaQXU9tJHBcAmKRhw1OvNLvdPgtZ7qHy47qPzIW3qdy8c8HjqOteiaOuh6jpk3hy21e4vlm+aMS2zEw47gkYUf1PvWvLommG+0KGSd5pLC3Jhj8rcJVXaCxx6Hbj39a55YzllaS79Ht0NVh7q6Z5Jf6VfaWYRe2skBmQSR7x1B/r7dR3qnXa+JV0K+uby7uNcvJNQCnZBJasgUj7qYI+UVmeF9As9aj1CW9upLeGziErMi5+Xknj6Ct41v3fPP8AJmbp+9yxMfTbCbVNQhsrcoJZm2qXOBnFGo2E2mahNZXG0ywttbacjNdr4d0/wtF4gsnsdauZrlZMxxtbkBjjpnFHiLT/AAtL4gvXvtauYblpMyRrbkhTjpnFR9Z/ectna3Zley9y9/xOAorovFnh238PzWYtrl547mLzAXXBFYYtbhrVrpbeU26ttaUIdgPHBPTPI/OuiFSM4qS2ZlKLi7Ms22j315pt1qEEBe2tv9awIyOnbr3z6YBqhXV6JdX3gy9gudQtidP1CEkojI/mpjIIwfcdexNYP2ZtS1CddLs5jH80iQrl2RBzyfapjNuTvt0Y3FWXcpU+GKS4njhiXdJIwRV9STgUyvRr1PCXhTVLPzdNu5LpY0uEZH3KDk44LeopVavJZJXbCEObW9jhNS0y80i7NrfQ+VMFDbdwbg/Qmqldxf674M1O+lvLvS9SknlOWbeB2wOj+gqbR4fBWtarDp9vpd+ks27a0kvyjCluze1ZrESjG84vz/q5fsk3aLRwNbUfhLX5YkkTS7go6hlOAMg1Ul024n1G7hsLSeZYpWG2JGfaMkDOPpXqUmp2skduBBuKQqjedY3O4EDkcJiivXlC3Ir3CnTUr8x51/wh3iH/AKBU/wCn+NZ+oaVfaVKkd9ayQO67lDjqK9R+3W3/AD7Q/wDgDdf/ABFc543EurPYNYWd1KlvAVlKWsqqhz/tKOKzpYmpKaUloXOlFRumcpZaJqepW8k9lZTXEcZwxjXOD1xjrT/+Ee1v/oD6h/4Cv/hUFrqmoWKMlnfXVujHJWGZkBPrwa67wUuu67qyyzarqP2G3IaVjcPhz2Tr37+31FbVZzppydrIzhGMmlrc5j/hHtb/AOgPqH/gK/8AhUd1o2pWNqLm7sbi3hL+WGmjKZbGcYPPQGt/xBrfiC51PULy3uL6CxguDADDIyxqRwBxxk4z+P0rnbrVdRvohFd391cRhtwSaZnAPrgnryadOVSSTdhSUForlSiiitzM9W8Bf8ixH/11f+dFHgL/AJFiP/rq/wDOivmsR/Gl6s9al8COD8W/8jXqP/XX+grFra8W/wDI16j/ANdf6CsWvoKH8KPojzKnxv1Ctbw9oVxr+qx2sQIiBDTSdkTufr6VlLjcNwJXPIBwa7lb177Sn0Lwfpk5heNTd3D4DvuHKk9B6dexwMc0q05RVo/f0Q6cU3qa3ipvDIh0+5ure8urPyvKtpLOQeUoXjb1AB/w9qxtS03w3F4TbVrexvYZJm8u1WeXlz/ewM8Dk89ce4rQ061tvDelT6b4p1C1ktZyCtimZJEb+9kcqOD/AEPrB4i8Oarr8y3mlXNpfafGuy2ht3CCJOm0KeM+pz26DAFcNNqLUeZ2vvd2f9dTokm03bXt1OP0bV59D1Fb62jheZVIXzU3AZ7j0P8Aia2LTSdN1m3Go6l4oit7y4Zmlikh3EHce+4devTvWRBol3LrY0mUxW1ySVbznwqkDPJGa2/+EA1D/oJaV/4En/4muurKmnfms318jCCk1a10dRrOj6XceEtHtZtdigt4f9VctDkS8dhnj864HXNL0/TTALHV49REgbeUj27MYx3PXJ/Ku61jw9Lf+FdJ0yK/08T2n+sLT4U8djiuabwDfqrMdS0rABP/AB8H/wCJrnw9SMVrPq9P6RrVi3tEb4EtLW71a9+120VwkVk8qpKuRuDLz/OpF8brb20sVjoWn2pkZGJVSQSjblyO+DUvw7SP+0NTlluIYYxYujGRwCASPm+gxz9RVqHS7G9tLrSfDWnG9ZsJcancvtVcHcNn5dvQdRVVHD2sudX29EKKlyLl8xt7YjxnZadq9tCFvGuBa6gsQPGSMPz2A+uAQP4ax/G98t94ru/LIMUGIFwP7vB/8ezXRWGpaP4C3WyTtqV9OwFy0DYSJQeg9WGTx+eKx9c8JMLZ9a0e5W+0twZC5f54/Xdnr/P1FFKSjPX4fs/P+tAmm46b9SLRLyfUNGl0NtSsNMs1PmPJINrzEn7pOef/AKwHSuk1LQobbwtZ6Bd6xY27RzG4WaRyN4O7oMf7XXPauL0/wvrWq2y3FlYvLC+dr71UHBx3I9K7BLG1sdEbTPGF9ZlIgPsvkyF7mEHqBgfd4Hr09AMKtyqXuy67Le4U7tar5kuo6Gms6Vp9xa6zZxx6RbiJ7mNywyoHORjaeM9+tcNrOu3uvTQS3vlGSGIRhkjClvcnuf09AK7jW9IvNc02K08NXFg+jwLxbwzYdm6/Pnv9T1OTXCapomo6K8aahb+S0gJQb1bIH0JqsK4vd69F1Qq1+i0M+iu706LQrn4c3j+RbtqVrEfNkaPDKWdtnzEYJIx0NcJXTTqc7atazsZSjypeYVYsb2fTr6G8tmCzRNuQkZGfpVerulaidK1KG8FvDP5ZyY5kDA/n0PoRVy2elyVuei+E5PFerSpe6hdmLTgN4Bt4g0/sPl4Hv+XqI9Rn8e3GoTSWdulvbFv3URe2YqPqSTnv+NQ+HdXvfE9v4h/tG7SJGtAi5+WKEENk4/maw/8AhENP/wChr0r/AL+D/GvMUYqpLnSXla513biuW/3lrVtT8caVZebqMqxQSnytwSBskg8fKCegNcTXeeJbJLD4e6Vbw3kN3Gl22JoTlWzvPH05Fc/4cvdH06aa71O1luZ4V3W0QI2M3+16eueR14ziuqjJKm5Rj16K1zGonzJN/edl4L0KPTNNlk1AmO+1KJlghVwspjAydvox6+2B0rK0+18Janqiafb6Rq/2hm2kM4GzHUt83GKdJpHinWNXGu3jR6WIsMks7hBCg5xjrxz1xnNbc13peqm8j0PU7KDXZ0WKa5MbJ5wA+byyTxkjqM9B7EcspNScr3b7bL1t0N0lZK23fqef+JIdOttcuLbS1YW0J8vczltzD7x/Pj8M96tW2qy6xbWGg3l5Bp+mxEkyrF3AJBbnk5+nLZ5qHVfCesaNC895agW6kDzVkUqc9O+f0qfTPCF5qunx3kV9p8aSZwk0xVhgkcjHtXY5U/ZpuV7dd9TntPmat8jrPB2h6Tp+vrPZ+IYb6by2HkpDtOPXO41jXvhrQXvLiRvFlujtIzFDb8g5PH3q1fCPhmfQtdW9u7/TmiEbJiOfJyfqBWRd+Bb6e9nmTUdLCySM4zcHOCc+lcsZr2rftOi10/yNnF8iXKcbXfand6ZoOk6J/wASGwunurKOV3kTB3bRk/jmuW1rQZ9DaFZ7m1mMoJH2eQvtx68D1rt9Zt9Aj0/RH1a5N1Jb2ccUdnaPkzEgDORzt44PH9K3rTjJwe6d9jOnFrm7lDT/ABkmt6tJp+qWVslhqLBZNoO4PtCqSe/IH04OeKiFhN4W8Ka+tzlZ7m4WziyvDqOSw9ipP5VNdeGo571tZ19bbRLD5QtpG+XcKMADHfA7D8O9WbvUNO+IEP8AZ8MsljeW5JtFnfKzDA6/7XHue4zzWF4prkXu6X7K39amln9rfocFp19JpmoQXsUcUkkLblWVdy59xXoGkaXNe6+niKXXLG8mhHmzRRscRgqRtyM4A+nauIufD+p2urjSntibxhlI1YHcOeR+RroNF8L+J9IvBqK/Z9PEQ+Z7qYBGXuGAzxW+IcJRupJNr70Z0uZOzRs6Vb6fa+I9R1BNb02dtREkYgWUgqZHB4ODn06c1h6gNQ8CCXTotQtrpb2JhLA0e4J0AbB4yRn2OOQeK6S1/wCEWh1O5n0W5sE1coFiErEQI57pxjJzjj8Mc1y+p+CvE73Mt1cRJdM/zvOs64P/AH0Qf0xWFKUXP33ZaaPr/wAMaTT5fdWvkT/DZWOs37hTtFi4JxwCWXH8j+VXpFW00HwTbO37x7sTDjHylw3/ALOK57wx4c/tuZ557iGCwtiGuXeQBgvXgds46nj64xWzdeJNA1bxDJHqNo76UsK29vICwMQBzvCj1OPfAA9RWlWN6ra17/c1+pMHaGuhsad/yU3xB0ObJjx/2zrAvv8AklGm/wDX+f5SVnieTwz4iuItEvoLxZU8lZcAgqxBxnpngcg4/kNnXtPutL+Gmn2l7F5Vwt8SULA8ESHtU8vLKOu9vwTHe6fzF0fxFoMmn6PpM2n3XmwTIysjgL5xP3zzk8mrniXUv7O1+X+zFutPuri4SK5vnUGJlx2zxx17dDWVofjC5SfS9Mt9PsIo90UDSeVl2yQC2c4yevTrS+L9dnHi2W0vgbvTrWdXW1JCgnb6gZ7ml7J+2tbo3vfqPnXJv+B1F/e3Wq6qqaP4ss4UYBI4FAdmPc1yfiKfXLfxBY6bcatHfzwyxyxkKAEkJ4Bxz6H8a2tEh0hrO/v/AAorf2u8RWO1uZBugHcqDnPbkk+nGSK4CC1nv9WS0lkEdzPNsZ5yeHJx83BOc08PTipPsu6V/wCvzFUk7Lz89D1vQpvFkmryjWYoI7VYnCmIphnyMdCW6Z/rXJavd+NbXQrz+1WjNrNtjcsYmYA56bfXv+GK1PD+i6f4S1SW5vvEFiWaBo/KDAMMkHPXPb0rl77T/Clpp83kaxdXl6q4RUh2IW/EdPxqKUY+00Sa0+y/6RU2+XV9+p1t1/yWOx/64H/0U9YnhSWyj8Tat5ssEN+xdbKWcZVXLEfn0/UVf8Mx61r/AIrsvEl3bxLbKjoZUIAOFZemSc5NctbW2j6hqWorf6k1mzTZt5RHvRssc7se2MHIqoxVnBvaKTtr1Ym3dSXdmj4lu/FVpHbxa0AwiuBLBchFPzDpgjjHfBGa6DXPE2r2fg7Rb+C72XNz/rX8pDu49CMD8Kw9Y1PTrDwl/YFrqT6pM0wkMxUhIlGMKufp0HqenSpob3Rte8I2Gn6hqY0+fT3JbMRbzE5+774x+IPHSm4JxjKUdE+3T0EpNNpPp3/UveFNLubbwfJqWmLb/wBqXUhSOaYgeUgODjPGcg/mPSp9Ts7668IagfEklnJdWw8y1uUdd/XleAOuAPfPtXPaNqumX3huTw7rExtVWTzba6CbgrHsR+J59CelU9S0nQtPsZjDrv268+Xyo4oSq9edx5HTPHFHs26j5t79unqHMuTTa3f9DoYdUfRvhrpd5BDE9158kcMki7vKJZ8sAe+Bj8fwpv2qXxX4C1C61PZJe6fJuiuAgU44JHGPcfl3FTW0Ol3Hwz02DVbh7ZJLhxFOqlhHJufBI7jGf/rVmX2o6XoXha40PS7039xdybp7hU2qq8cDPXgY79T06Ukk5PlXvc2/lfuNtpavSxg6Bolxr2qR2kCkJkGWTtGnc/4Vq+Odbt9V1WG3smD2llH5SOOjHuQe44A/Cty1vvCsPhZdIi1uS2aZQbqVLZ98hPUZ29O2OeOPXPKa3ZaDbQRPpGrSXkhbDo8DJgY65IH+TW0Jc9XmkmrbaP7zOUeWFk156mLRXd+CotCutB1CC9gt5L+MSzgyR5KxBF53YwADnvXCVvCpzSlG2xnKFkn3CnRyPDKksZw6MGU46EU2pbW4NrdRziOKQo2dkqB0b2IPUVo9iUeg+G7zxh4iulma98jT9xLzfZoucfwrleT2z2/StHXZ/Gk2puNIt1gskG1Az27F8fxfMTjPpVLwp4mv/EHiplnZYrZLaTy7ePhF5HJ9Tjv/ACrB/wCEQ0//AKGvSv8Av4P8a8vlSqPnSVulr/kdd3ye62/nY0L6/wDH2l2T3t3IscEZG5wtu2MnA4AJ6mua0KxbW9fjjmhmuEdjJOISqttzyRnjqa6u50yLTPhtqcNvqNvfKbpHMkDZA5QYPv0P41xWlQalcXoi0sTm5Kn/AFDEHHfkdq6aLThJxsul7WMp3Ule7/E7y9ttet7J9O8O+HpNPtX4eUyKZpPqd3Hfufwpdc0TX2sPDj6dbyrdWVtsfy3AMbYUdc85wapWei6np6rfeJtbnsrVTnyPtTGSXH8Iwe/tk/Sl/wCE0g1y7uLHUZZ7GzlkBtLmB9rwYGMNjqp6nrgn6Ec6U73hZpbvX076mt42tLS/9fIl1qz1rWNAxqPh0nVIsf6arouUHJJAPJ9unOeOlUfAMC3Fn4gjkmWGN7Io0jdEBDfMfYUzVdA8S2Vq93Bqct/p+wuZ4bo42Yzkgn09M1yCTSxK6xyuiuMMFYgMPQ+tb04c9Jxi18un3mcpcs02jvfD3hvTbPxBZXEPiSyuZEkysKD5n46DmjxD4b0288QXtxN4ksraR5MtC4+ZOOh5rm/BwJ8X6bj/AJ6/0NHjH/kbtT/67f0FLkn7e3N08u4c0fZ7dTe+JKhJdHRWDhbXAYdD05puqs2m/C/S7MsokvZTKy55KZLD/wBkriXlkk2+Y7PtUKu45wB2HtV7UdavdVtrOC7dGS0QpFtQLheOOOOwrRUGowjfRO5LqJuT7nReNP8AkB+Fv+vBf/QUrB8P6zNoOsQ30XKg7ZU/voeo/wAPcCu11fw3f+I/Dvh2TTjCywWSo+58c7V/wNU7XwWmj6Nql54hihZY4s24SYg7xnAyPU4Hf6VlTrUlS5JavXT5lyhNz5l/WhjeNLPTrfWRc6ZcxSQXaecY0bJjY88jsDkED6+1dTqSJJ8TdDR1VlNouQwyDw9eZV2/jm9uNP8AFVhd2snlzx2UZR8A45Yd6qdN3jC/Rr8hRktZW6otv4tuV8WtpQ0/Tfs4v/s2fs/zbfM29c9ce1OgRI/jNsjRUUE4VRgD9wa53/hOvEv/AEE2/wC/Sf8AxNb/AIO03VtS1+28UXc0UkJMnmyFgGyEKcgDA7fhWc6fsoOUrLRrfd/cXGfPJJa63OQn1C9sNWvjZ3dxbl5nDGGQpu+Y9cGvTNR8mwlhj+2gF4lc/atdnhbJ/wBnnj3rzm202fW/ENxFaQvcKZWkcRMoJj3ckFjjvXpAj1dAFSPxHtAwP31p/hRinG8dfxsKjfUjvngtPs2L6IedAsv77xBOmc5+713L6HjPpWB42uLnToNMax1G8RLqFmkCX8sqN06Enkcn612GoLqf+i+SmtH/AEdN/kS2w+bnO7cPveuOPSuQ8XaPqt9ZJdtbao/2ZWaRryWAqqYycbMHPFYYdrni5P8AE1qp8rsvwMHw94Vn1pWu5pUtdNiP764c44HXH+J4/lXVaJ4gtbjxVpOjaKjwaVAZc54M7eW3zHvjvz+I4GPOVuJ1t2t1mkELkM0YY7WI6EjpXQeAf+R20/8A7af+i2rtxFNuMpSeydl8jnpTtJJFzT/Eo0bX9Us72IT6Xc3MgniIzjLEbh+Haqfizw7Bo08FzY3CTafdjfAd4LD29xyOfz98jWf+Q5qH/XzJ/wChGqjOzBQzEhRhQT0Gc4H4k/nVwpWkpxdu/mTKd04sbRRRXQZHq3gL/kWI/wDrq/8AOijwF/yLEf8A11f+dFfNYj+NL1Z61L4EcH4t/wCRr1H/AK6/0FYtbXi3/ka9R/66/wBBWLX0FD+FH0R5lT436hVi1v7yxEgtLqaDzRtfynK7h+FV6K0aT0ZN7CsxZizElickk8mpLe6uLSXzbaeWGTGN0blTj6ioqKLCHSSPLI0kjs7uSzMxyST1JNNoopgFFFFABUsV1cQJKkM8saSrtkVHIDj0OOoqKila4BUq3M6W726zSLC5BeMOQrEdMjoaiooAswajfWsDQW95cRQsdzRxysqk+uAarUUUWSHcfFLJBIskUjxyL0ZGwR+NPuLq4vJBJc3Es7gbQ0rliB6ZNQ0UWV7hcuf2pejSv7ME22zL+Y0aoBub1JAyfxPYelU6sWtheX7MtnaT3DKMsIYy5A98Va/4R7W/+gPqH/gK/wDhU80I6bDtJmbRWl/wj2t/9AfUP/AV/wDCoLrStRsYhLd2F1bxltoeaFkBPpkjrwaanF6JhyvsRRXVxbxyxwzyxpKNsio5AcehA69TUNPihlnlEcMbySHoqKST+Aqxp+mXmq3DQWMDTSqhcqpGcD6/WhtLVis3oRNdXD2yWzTytbodyxFztU+oHTufzqGrCWN1JYyXqW8jWsThHlC/KpPQE/56j1FV6at0B36lm61G9vgou7uecIAFEkhYDAxxmq1SrbztbtcLDIYEO1pAp2g+hPSr9loN3f6ReanC0XkWn+sDMQ34DFTeMV2HZtlKS9u5bZLaS6meBDlImkJVe3A6CoKKKpJLYQUUVpalol5pVrZXFz5fl3iGSIo+7jjr27ik5JNJ9Qs2rmbRRRVCJbi6uLtw9zPLM6qFDSOWIA6DntUasyMGUlWByCDgg1Zl068h0+C/kt3FpOSscuPlJBII9jwevoaWbTLy30+C/lh22s5Iik3AhiOvQ+xqU47DsyJru5e6F09xK1wGDCUuS+R0OeuRilub26vWDXVzNOw6GWQsR+dQUU7ILsKsR6hexW7W8d3cJAwIaNZCFIPUYziq9FDSe4XFDMoYBiAwwwB6jr/QUlFaC6NePob6wEX7IkvlElsHPHQdxz2pNpbgk3sZ9WLq/vL4qbu7nuCn3fOkL7fpmn6Zp02q6jDY25QTTEhS5wOmf6VHe2kthez2k4xLC5RsdMg9vai8ea3ULO1yJHeORZI2ZHUhlZTggjoRTpppbiVpZ5Xlkb7zuxYn6k1HRTsIkhnmtplmgleKVfuvGxVh9CKjq7c6TfWlhb308BS2uP8AVPuB3fgDmqVJNPVDaa3Ciiruo6Rf6S0S31s8BmTem7B3D8O/t1p3Sdgs9yNNQvY7VrWO8uEt2zuhWVghz14ziq1aOnaJe6pa3lxaIjJaKGl3OFwDnnnj+E1nUk43aQO9tQoq7aaVeX1ld3dvEHhtAGmO4DaDnB56/dPSqiI8jqiKWdiAqqMkk9hTumFmNorT1vQb7QLpLe+RQZE3oyHKkfX1rMpRkpK62Bpp2ZrXGuyXHhq00UwoI7eUyiQE5Yndxj/gVZNFXH0u8j0qPU2hxZySeUsm9eW54xnPY9qSUYeVxtuRTooqxY2NzqV5HaWkRlnkztQEDOASevsDVNpK7EtSS31S9tLG4sreby4LnAmCoNzgdt2M49s45PrVOug/4QnxH/0C5P8Avtf8aqXfhnWrJlWfTbgMwyAi7/8A0HOKzjUpX0aKcZ21TMqirU2m39vE0s1jcxxr1d4mAH4kVXRGkdURSzMcKoGST6VomnsTZklvdXFpIZLaeWFyNpaNypI9MioauS6VfQakmnTWzR3bsqLG+Bkt056fjTm0fUE1f+yjbMb7eE8pSDyRnqDjpznPFTzR3uOzK63Vwls9ss8q27nc0Qc7WPqR07D8qS2uZ7OdZ7aV4plztdDgjIwcH6Gi5tp7K5ktrmJopo22ujDkGp7fS766sbi9gtpJLa3x5sijhc/559O9NuNrvZi1uQT3E91KZbiaSaQ9XkYsfzNRUVoaho93plvZT3IQJeRCWLa2TtIB59OooulZBZvUrR3l1FbyW8dzMkEn341chW+o6HpUFFbVt4R167to7iDTpHhlUMjBl5B/GlKUYat2Gk5bGTDNLbyrLBK8Ui/ddGKkfQiklmlnlaWaR5JGOWd2JJ+pNbv/AAhPiP8A6Bcn/fa/41T1Hw7q2kW63F/ZPBEz7AxZTlsE44PsalVabejVxuE0tUZdFFFakBRRRQAU+WaWdg0sjyMBgF2JOKZRSAKkE8wgMAlkEJO4xhjtJ9cfgPyqOigAooopgFFFFABT4ZpbeVZYZHjkXo6MQR+IplFIBWZndndizMckk5JNJRRTAKKKKAPVvAX/ACLEf/XV/wCdFHgL/kWI/wDrq/8AOivmsR/Gl6s9al8COD8W/wDI16j/ANdf6CsWtrxb/wAjXqP/AF1/oKxa+gofwo+iPMqfG/UKKKK1ICiiigAooooAKKKKACiiigAooooAKKKKACiiigC9pM96l/FBZXs1o9xIsReKRl6nHODz1ruBpuoWOpxwXXjj97HIu6B7mTJ6HGM9/wCtcPo6t/bWnNg7TdRgHHBO4f4iu61288Op4zkiutLuZb7zowZlmwu7C4OM9uK4sRfnsu3ZfqdFK3Ldl7xVDPJrbMni1dMXy1/0czsn44BHWuK8WW+q6ZfLpuoatcXy7VmG+VmUE5HQnr1/Oup8Z3fh2HxCyanplzcXHlqfMjm2jHYYzWR8SlZvFahQSfsqnAHYFiaywracE9rdl5F1rWk/Md8ObJG1K61JpYQ9rERGkhx87AgHPpgMK3PDvhu+0w6vqU97Yy3c8LRxTLMSodjkljgY5xXmdtbTXlzHbW8bSTSNtRF6k16jLpek6V4NbTrwTXRtGWa+is5BvDsOrdMqO2ewB7U8UnGW/wAVtLdBUXdbbGfoGh6vobzRm80e6srhCs1rJdHa3HX7vB/mPwxxuuaNLo1ykc09pIZQXC20hcKM9Oea6vStO8IarBd3CWGpQ21rGZJZpZBtHsMEkn2rhvORLzzooVCLJvWKT5xjOQp9R2961oczm39+lv1IqWUUj0Sz0pLHwCunXGqW+nXGpN57m4fadnHygcHoFz9TVjRdCtrbwlrFmmtWU8c+N1wjfJFx/FXLLOnjG9mutb1qDTmhVEhTycqRznA3DHTJ69a6vRtH0u38Jaxaw67FPbzf625WHAi47jPP51zVU4RtJ6tpvT9bG0LSei0t3OXPhHTlVmPirSzgEgK4JP61zunXRsr+KdbWC6ZSQIbiPejZBHK9+vHvitjVNB0ex0+S4tPEcN5MpG2BYdpbJAPO49OtY+nX0mmahBexRxSSQtuVZV3Ln3Fd1NuUXrf1Vjnlo1pb8T0eGPyNa0ix1DR/Doa+TdJDFZ4li4J55IxwBn68cZp+n6yNT1zVdFudL0xrLTUne3T7PnBRwo4Jx0J6AVV0jS5r3X08RS65Y3k0I82aKNjiMFSNuRnAH07U7SrfT7XxHqOoJremztqIkjECykFTI4PBwc+nTmvPko69Xbz3/wCGOlX09fwM+8vLy00O01hfD3hue0nX5mis8+Uc8Buf8nj0zyGk6XcazqcNjbKS8jctjIRe7H2FdHqA1DwIJdOi1C2ulvYmEsDR7gnQBsHjJGfY45B4rJ0fxJcaJYXtvawxCa5AAuMfOg74/wA9a7KakoN09b7f15fiYTtzJSNr4hyG0u7DRYfls7S2UouTyTkZPrwP1PrSXttPefD7w3bW0TSzSXEqoijkne9aHi7QNR17xdHFZQ5VbSPfK/CJyep/p1q/ql9/wh/gywg06WK6m3SQx3WAQhLEuQPrkfh3rnU/cpxjrK9/we5o4+9Jvb/hjlPEOkaVoWmW1gztNrmd87Rv8kYP8JHr6Ywe57A8xTpJHlkaSR2eRyWZmOSxPUk+tNr0KcXGNm7s5pNN6IKKKKskfDDJcTxwxKXkkYIijuScAV6H4stn8N+FtCtrWUCW2uBIXAyPNAznn3Jrg9N1CfStRgvrYr50Lbl3LkH1BH0rvMW2veG9EfWroww3F1cySzeaqbTliOWzxnAx/hXHiW1OLfwr/Jm9Kzi0tzRGoXcVlYy3fiHUllubaO4Kw2Cuo3DpkJ9abcX9+dHv72x8Q37SWkYcrPYogPPuoz0NQWuqSLaQi1h8X/ZxGoi8u1hK7McY+Tpilubw3lu9tfJ4tS1kG2V5reJEVe5Y7OAOpNcfJrsvw/yOjm03/P8AzPN7y8n1C8lu7qTzJ5TudsAZP0HFX9W0KTSbDTLt51kW/h81VC4KcKcH/vqoNatrOz1i4t9Pn8+1RgI5d4bcMDuODzmuj8Zf8i54V/68z/6Clei52lBR2f8AkcijpK+6LesaTqWq+D/DSafbSzrHC5kCdATtxn9f1rlb7w9q+mW/2i8sJoYdwXew4ya6TXtRvrDwl4Y+x3lxbb4JN/kysm7GzGcHnqa5W51fU7yEw3Wo3c8ROSks7MufoTWeH9pbS1rv13ZdXlv56fkWdDt9EuHmGs301qqgeWYkLbjznPB9q77UY/Dtx4L06O71SdrBJClvcmI+ZlcjbwOgAI5HYVxmlaFpF9p6XF34jhspmJBgaHcVweOdw611t7omkv4K0yyk8Qwx2yTO8d0YeJDlsgDdxjJ79qyxEoua1e/b8tC6Sai9Ft/XUksLfw3D4K1GK11W6Fg8wFxceWQ4J2jaBt6Hjt3NcHrlvolu8I0a+mulYHzDKhXaeMY4HvXcWeiaTH4J1KzTxBDJayToz3Yh+WMgrgY3c5wO/euS1XQtIsdPe4tPEcN7MpAECw7S2TzzuPSjDyipv3nv28uugVU+VaLb+up09tNp0XgPfB4fmmhvJRvto7hyXCYzISBlQGHb29aNL0vSdGv01u92aaJx/wAS+0uHLGM7OXfv1z9MjvjFvS9a0C6ilt7Vb949P0iVWB2qrR/KXwRzvJA54HWq2g3jP4TuJbbV00uJdTZIJLoCXbFsBWMbvz/A1i+a0lqrvz6/f+RorXXUUTaNPpc1jrviK31GOSXfFKilZIXY8kdeOenbnt04vX9DGh3MSJf295FMnmRvCeSvqR2zzjk9K7jWoIZ77xhLLDG0tpBbfZ3KDdFlCTtPY5rlPHoA8a6gAMD930/65rW2Gk+eye6v07L/ADM6y93X+t/8iz4rjjTw34YZEVS1qxYgYJOE60t1/wAkrsf+wkf/AEF6u6pr+u+EnttIE9pMkVupRvI6LyAOfpWjP4u1SPwJbaqvkfaZL0wtmIbdu1j09eBSvNRjZXV9NfXyC0bu76HPeGYYpfCnihpI0dkhiKllBKnL9K5aOWSGQSROyOOjKcEV2drr/iLxZBe6YktlGn2dpZSY9uUBGQCAeea5rRptOg1FW1S0+02pBVl3smD2OV5ropuSc3Ja9l6fIyla0Uj1PVbS4k8RPIugzXEZdP8ASRqvlA8Dny89vTvj3rMutYS38X6xp8uq3VlJO0QtpEQOiNsHUHPUn2HXmrt3JZyXf268XwzFM7jE0k5mZiuAR91eQAB7VNqeoaXYX0ep397aIbiPNnJHaNIyLjkh+hJyOcD8a8yPRNX0t+Xqdj7p9f67GF4yl1XTPCkGn6hcPdz3U5aWcLhAqn5VGAOTgH8DXJ6Ha6JMkkmqarNZyo48pYoixPvkA10OtX2o6VYvdWNxcXuj6kpUtqEYJ8w8llBA4wOOMdeOlZnhHQ4dS+03N3pl3fQxAKiQSKoL9wckHoc9fWu2n7lFtu3p/k9jnn71RW/E7a0t9K8SxafNFqMtw2lyoftTwFWkIIwpZuvvjnp+M7WlhZ6nrerRXM4uWIhllS2MhtztH3cDnggng9qw5YPFN9qenRf2N9h0q2uY3FtC67QAwOWIPP5fhSanp/iuy8Y3upaPbyNFKykkFdkgCjggn6j19K5eS7tzrbuu60ubc2l+U5fVrHRpPK/sjVLnUL6ecKySxFS27POSOSTj866q58P3tvpWl+FrW3m23Mgm1C6SM7BznG7GOMf+Or61k+LLO98611aHQX0yVSDNIrqQ0pPGAD698ZOfaukg0vxPbeHJlM8txql8SHaW4+W1jwenP3v93pn2repU9yL5vvtv8tNDOMfeat/X/BOTlPhLSp5rGezutQkhkZTcpIYw3PTGe3TPfGa3vEl34djsdEN7plzNG9kht1SbaY0wMA88npzWZodl4j099T06x1OwtEsmWS5klwyEsvXJQ9APatKCfxTdzpBB4t0KaVuERGRifoPLon8SfNt5vt6aBHa1t/Jf5nDavNpk92j6VaS20AjAZJH3Etk5Oc+mK7q//s3/AIRjw7/aGq3dgfsvyfZ1J38LnOPTj86457aXU/FM1tquoQwzNK6zXTgBAyg89upGO3Wu+05maG00y28QaJdGJBHCjW29iAP9/wBqvENJR121/p2ZNJXbIL7+x/8AhENKMmuX8dr5knl3Ko2+Q7jkHvxz+VZvib7N/wAK90/7HeTXcH204mmBDMcP1zXayW91Np9rBFd6YbhWkDMbXcjYYghV3cEdDz1BrldYWDWtPSyn8T6KkKSeaqwxbPmAI/v+5rlozTkvJ3/PyNpx0+X9dTzaiiivZOAKKKKACiiigAooooAKKKKACiiigAooooAKKKKACiiigD1bwF/yLEf/AF1f+dFHgL/kWI/+ur/zor5rEfxperPWpfAjg/Fv/I16j/11/oKxa2vFv/I16j/11/oKxa+gofwo+iPMqfG/UKKKK1ICiiigAooooAKKKKACiiigAooooAKKKKACiiigDofDHiHVNNu7WwtLry7Wa6QyJ5anOSAeSM9AK3fFni/XdM8T3tnZ33l28ZXYnkocZQHqVz1NcGjvFIskbsjoQyspwQR0INOnuJrqZpriaSaVvvPIxZj9Sa55YeEqnO0vuNVVko8qZ1en+K/Geq3Jt7G7aaULuKrDEOMgZ5X3FavxB8RapZatJpltdbLSW2Akj8tTndkHkjI4964G2u7mym821uJYJMY3xOVOPqKbNPLcTNLPK8sjfed2LE/Ump+rQ9opWVl5D9rLltd3Or8OataabYomkadNdeILgtHlxuEYx1XHb29jk4AzpaJpWraBfvrGtX9tYwTFjcRztvafPJG0dSeTxz7HkVwdvcTWlwlxbyvFNGcq6HBBonuZ7qUy3E0k0h6vIxY/mac6Dbdno9+rCNSyV+h6JqGnQ+IdHSz8J3lpHZq5kmszlJHck8sTyRxwDgcewxwuqaLqGjSRx6hbGBpASoLA5x9CaoqzI6ujFWU5DA4INS3F3c3kgkuriWdwu0NK5YgemT2qqVKVN2Tuvx+8mc1LW2p03/CAagRkalpXP/Twf/ia6TR/D0th4V1bTJb/AE8z3f8Aqys+VHHc4ry+ipnRqTVnL8Co1Ixd0vxOs/4QDUP+glpX/gSf/iaxJtFu49abSofLubkHA8hsq3GeCcdqzqdHI8MqSxOySIwZWU4KkdCD61pGNRbyv8iG4vZHY6L4X8T6ReDUV+z6eIh8z3UwCMvcMBniuhtf+EWh1O5n0W5sE1coFiErEQI57pxjJzjj8Mc15lc3t1esGurmadh0MshYj86grGWHlU1nL7v17lqqo6JHWan4K8Tvcy3VxEl0z/O86zrg/wDfRB/TFcnViPUL2K3a3ju7hIGBDRrIQpB6jGcVXremppWk18iJOL2O2+IN5c23iAxQTyRxz2UaSqrEB1y3B/z61ah0O78RfD3R4LAxNLBLKZFd8Yy7f/WrhJ7ie5cPcTSSuBgNIxY49OairJUGoRinZov2icm2tGdpN4Nh0Xw5qF3rkqJdfKLRYpOS3PHvn6cBTU/gqLQrrQdQgvYLeS/jEs4MkeSsQRed2MAA571wlXLfVL20sbiyt5vLgucCYKg3OB23Yzj2zjk+tEqM5Qactb+gKpFSukU6KKK6TEK6zVAR8NtC97ib/wBCauTqRriZ4EgaaRoUJKRliVXPoO1ZzhzNPs7lRlZPzOn8JN4g1HVrMWt1dy2lpND5yfaSFSPd02k8jAPAqHxZq98fEep28Wo3P2YyNGY1mbZjoRjOMdeKwbe9urPd9luZoN4w3lSFd31xUFQqP7zndiuf3eU1tE0eDV3mWfVLSw8sAg3Lhd+c9MkV2mt6TpmraZpFoniXSo2sIPKZjOpDnCjI5/2a81ooqUpSkpKVrBGaStY9I1bR9N1LRtIsV8S6UjWEbozGdSH3benPtWFN4Pso4JJE8UaQ7KpIQTD5iB061ylFTCjOKsp/ghyqRlq4m54d8L3niKVjEyQ20ZxLO5B2/hnJ/l71b8ZavaXk9rpmmNnT9Pj8uNgeHbuc9xwOfr61zkc0sIcRSunmLsfaxG5fQ+o4HFR1fs26nPJ7bE8yUeVHYaaD/wAKu1k9jdx/zSuPqVbidbdrdZpBCxy0YY7SfUjp2FRVUIcrk+7uEpXt5Ho3hbWrGRb6DSdCgt7iHTpJfMZy5kcbflOedpY+vYVFY6pfX/hZ7ldPsb28u9X2CGeEMmfJHIBIAOF/nXCW93c2js9tcSws6lGMblSVPUHHb2pVvbpIo4kuZljjfzEQSEBW/vAdj71g8Krto0VZ2sz0jUZPO17xdpcKPJeXkMAhjUfe2x889B1Fcp4+/wCR11D/ALZ/+i1rBju7mG6+1RXEqXGSfNVyHyepz15yaZLLJPK0ssjSSMcsznJP1NVSw/s5J36W/L/IU6vMref+f+Z33izS7bX9WjvLXX9FjjECxkS3gDZGfQH1om0m2k8FW+jDX9E+0xXZnZjeDZtww64znkdq8+opLDyUVHm28g9qrt23PRfCWg/2Xd6hP/a2l3e6wkTy7S48xhypyRgccfqK5bwtHp02rCK/sp7xmX9xDE4G989DnHGM9x071k293c2js9tcSws6lGMblSVPUHHb2qNHaN1dGKspyrA4IPrVKjL3rvcTmtLLY9K8QXFhZWyrryxySrGVtdItHxHCCMBmIH3sdD25wO9PufEFvouk+GUvrGO7tJLNXIKBmjdVXDLnjPJrzJ3eR2d2LOxJZmOSSe5p8t1cTpGk08siRLtjV3JCD0GegrNYRWSb/ryK9u7to7XXNG/t2GbVrTxLFfQQp5jR3B2vCuP7qjg8dNozXPaJYeIL5JBo63ZjDfOYpSi59zkDNY1WbXULyySRLS6mgWXG/wApyu7GcZx9TWqpyjDlTv2uv8iHNOV2ju4GfwejXutavJeaiEIgsEuGYKTnl+en+eTiq9rq6eLLeOG51eXS9YjUIsiyFYbkdsgEAN9PXv0HCMxZizElickk8mkqPqy3b97v/wAAr2r2Wx02r6Lr9hJDDrN3KLKWZY/PacyR898E54GTzjpWldadoQwur+M577nKrCGfGPfLDv7Vx0t5dTwRwS3M0kMX+rjdyVT6DtUFV7KTSu7ei/4cXOlsvvPQPDUGlta+KobW7lTTTbxgXEqbmUbWySoxnnPFR+FtL0GDxLZS2niP7VOrHZD9hdN/yn+InArh47ieFJEimkRJBh1ViAw9D602OR4nDxuyOOjKcEVLw8nze89fTtbsNVUrabGzdWsN94zvLa4u0tIpLyUNO4yqfM3XkVtza1ovheCW38OBrnUHXY+oSDIUZ52j/AY6da4pmZ2LMSzE5JJySaStJUVKyk9F0JU7Xtudxc6jdWHgbw3qEEh+0R3Mzh25yS75z655oeXwr4pRri6kOjaltYybQDHIeu73PtkH69a4xrid7dIGmkMKHKxljtU+w6DrUVQsOt72d3qvNj9r5aBRRRXSZBRRRQAUUUUAFFFFABRRRQAUUUUAFFFFABRRRQAUUUUAereAv+RYj/66v/OijwF/yLEf/XV/50V81iP40vVnrUvgRwfi3/ka9R/66/0FYtetan4P0u+1O4uZvPMkjkth8D+VVP8AhBNG9Lj/AL+f/Wr1aWMhGEU09jjnQk5NnmFFen/8IJo3pcf9/P8A61H/AAgmjelx/wB/P/rVp9dp9mT9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkeYUV6f/wAIJo3pcf8Afz/61H/CCaN6XH/fz/61H12n2YfV5HmFFen/APCCaN6XH/fz/wCtR/wgmjelx/38/wDrUfXafZh9XkT+Av8AkWI/+ur/AM6K39C0a107Tvs1uZBGHJG5snmivErSUqkpLqzvpq0Uj//Z">

step1:选取K值

k 的选择一般是按照实际需求进行决定，或在实现算法时直接给定 k 值。这是基于项目你想要聚类的个数来决定的，但是也有不确定的情况，我们可能需要去一个最优的K值来将数据很好的归类达到最大化区分类别，这时候就需要思考从数据角度出发，应该进行怎么样的计算能够得到最优的K。

手肘法

手肘法是最常用的确定K-means算法K值的方法，所用到的衡量标准是SSE（sum of the squared errors，误差平方和）。

SEE各个计算在K-means里含义如下图：

误差平方和又称残差平方和，根据n个观察值拟合适当的模型后，余下未能拟合部份()称为残差，其中y平均表示n个观察值的平均值，所有n个残差平方之和称误差平方和。

在回归分析中通常用SSE表示，其大小用来表明函数拟合的好坏。将残差平方和除以自由度n-p-1(其中p为自变量个数)可以作为误差方差σ2的无偏估计，通常用来检验拟合的模型是否显著也用来寻找K值。


import matplotlib.pyplot as plt
import pandas as pd
from sklearn.metrics import silhouette_score
from sklearn.cluster import KMeans
data=df_true_data.iloc[:,1:]
distortions=[]#簇内误差平方和  SSE
for i in range(2,15):
    Kmeans_model=KMeans(n_clusters=i)
    predict_=Kmeans_model.fit_predict(data)
    distortions.append(Kmeans_model.inertia_)
    print("簇内误差平方和：",distortions)
#SSE  手肘法
plt.plot(range(2,15),distortions,marker='x')
plt.xlabel('Number of clusters')
plt.ylabel('Distortion')
plt.title('distortions')
plt.show()

簇内误差平方和： [95795.9293255097, 84414.58708674107, 74423.36972115413, 67317.01104077617, 60994.96414080261, 55045.26807434524, 50777.37404069404, 47753.98868450008, 44842.179122758884, 42867.62108059854, 41247.78745541904, 39800.985182972385, 38372.77525889944]

根据拟合图片我们知道选K为8时能够得到最效率的K值。

step2:计算初始化K点

初始质心随机选择即可，每一个质心为一个类。对剩余的每个样本点，计算它们到各个质心的欧式距离，并将其归入到相互间距离最小的质心所在的簇。


 
def euclDistance(x1, x2):
    return np.sqrt(sum((x2 - x1) ** 2))
def initCentroids(data, k):
    numSamples, dim = data.shape
    # k个质心，列数跟样本的列数一样
    centroids = np.zeros((k, dim))
    # 随机选出k个质心
    for i in range(k):
        # 随机选取一个样本的索引
        index = int(np.random.uniform(0, numSamples))
        # 作为初始化的质心
        centroids[i, :] = data[index, :]
    return centroids

step3:迭代计算重新划分

计算各个新簇的质心。
在所有样本点都划分完毕后，根据划分情况重新计算各个簇的质心所在位置，然后迭代计算各个样本点到各簇质心的距离，对所有样本点重新进行划分
重复2. 和 3.，直到质心不在发生变化时或者到达最大迭代次数时


# 传入数据集和k值
def kmeans(data, k):
    # 计算样本个数
    numSamples = data.shape[0]
    # 样本的属性，第一列保存该样本属于哪个簇，第二列保存该样本跟它所属簇的误差
    clusterData = np.array(np.zeros((numSamples, 2)))
    # 决定质心是否要改变的质量
    clusterChanged = True
    # 初始化质心
    centroids = initCentroids(data, k)
    while clusterChanged:
        clusterChanged = False
        # 循环每一个样本
        for i in range(numSamples):
            # 最小距离
            minDist = 100000.0
            # 定义样本所属的簇
            minIndex = 0
            # 循环计算每一个质心与该样本的距离
            for j in range(k):
                # 循环每一个质心和样本，计算距离
                distance = euclDistance(centroids[j, :], data[i, :])
                # 如果计算的距离小于最小距离，则更新最小距离
                if distance < minDist:
                    minDist = distance
                    # 更新最小距离
                    clusterData[i, 1] = minDist
                    # 更新样本所属的簇
                    minIndex = j
            # 如果样本的所属的簇发生了变化
            if clusterData[i, 0] != minIndex:
                # 质心要重新计算
                clusterChanged = True
                # 更新样本的簇
                clusterData[i, 0] = minIndex
        # 更新质心
        for j in range(k):
            # 获取第j个簇所有的样本所在的索引
            cluster_index = np.nonzero(clusterData[:, 0] == j)
            # 第j个簇所有的样本点
            pointsInCluster = data[cluster_index]
            # 计算质心
            centroids[j, :] = np.mean(pointsInCluster, axis=0)
    return centroids, clusterData

最后得到聚类结果。

五.画像分析

可以通过数据可视化横向和纵向对每个类别进行分析，由于我们之前做过特征方向标注，故进行类别分析的时候更好对每个类别用户画像区分：

比如分析聚类0：

BALANCE                     mean    545.247537
BALANCE_FREQUENCY           mean      0.827954
PURCHASES                   mean    462.690070
ONEOFF_PURCHASES            mean    212.797808
INSTALLMENTS_PURCHASES      mean    250.254275
CASH_ADVANCE                mean    207.813824
PURCHASES_FREQUENCY         mean      0.487740
ONEOFF_PURCHASES_FREQUENCY mean      0.141411
CASH_ADVANCE_TRX            mean      1.243806
PURCHASES_TRX               mean      9.470777
PAYMENTS                    mean    677.884833
PRC_FULL_PAYMENT            mean      0.176594
TENURE                      mean     11.435975

从均值进行分析，结合特征即可。

文章知识点与官方知识档案匹配，可进一步学习相关知识

算法技能树首页概览44726 人正在系统学习中

技术分享/学习交流/商业合作

微信名片

深圳幻海软件技术有限公司

一文速学数模-K-means聚类算法实战：信用卡用户画像聚类分析

前言

一、用户画像概述

1.用户画像

2.为何用聚类算法作用户画像

二、数据质量校验

1.数据背景

2.数据说明

三、数据预处理

1.数据空缺值检验

2.数据归一化

四、K-means聚类

step1:选取K值

手肘法

step2:计算初始化K点

step3:迭代计算重新划分

五.画像分析

我们是怎样将网站加载时间减少 24% 的？

程序员经典面试题，消息队列怎么用，才能保证万无一失

Python工程师需要掌握的面试题

是时候停止使用JavaScript IIFE！

SpringBoot实现全局异常处理总结

十行代码绘制漂亮金融K线图，这个神器你不可错过

Python 教你快速批量下载抖音去水印短视频

归并排序和快速排序

深圳幻海软件技术有限公司

一文速学数模-K-means聚类算法实战：信用卡用户画像聚类分析

前言

一、用户画像概述

1.用户画像

2.为何用聚类算法作用户画像

二、数据质量校验

1.数据背景

2.数据说明

三、数据预处理

1.数据空缺值检验

2.数据归一化

四、K-means聚类

step1:选取K值

手肘法

step2:计算初始化K点

step3:迭代计算重新划分

五.画像分析

【算法系列篇】滑动窗口

走进人工智能|深度学习 算法的创世纪

走进人工智能|深度学习算法的创世纪