带你用SmartNoteBook预测2022足球世界杯冠军球队

波澜 · 发表于 2023-3-10 12:02:45

原文链接：
https://www.kaggle.com/code/sivahemang/fifa-world-2022-analysis-and-prediction/notebook

2022卡塔尔足球世界杯

2022年卡塔尔世界杯（英语：FIFA World Cup Qatar 2022）是第二十二届世界杯足球赛，是历史上首次在卡塔尔和中东国家境内举行、也是第二次在亚洲举行的世界杯足球赛。除此之外，卡塔尔世界杯还是首次在北半球冬季举行、首次由从未进过世界杯的国家举办的世界杯足球赛。
揭幕战将于当地时间11月20日19时（北京时间11月21日0时）进行，由东道主卡塔尔对阵厄瓜多尔，决赛将于当地时间12月18日18时（北京时间12月18日23时）在卢塞尔球场进行。

世界杯是所有足球迷们的盛宴，更是所有球员们梦想的舞台，每支参赛队都怀着冠军梦想而来，但冠军队伍只有一支，32支球队64场比赛，谁将最终问鼎大力神杯？
我们看到已经有不少知名博主和自媒体根据经验做出自己的预测，本次小编也来凑凑热闹，通过分析历史数据并使用SmartNoteBook建模，实现从数据科学的角度来预测一下本届世界杯的冠军球队。

卡塔尔世界杯32强名单
A组：卡塔尔、厄瓜多尔、塞内加尔、荷兰
B组：英格兰、伊朗、美国、威尔士
C组：阿根廷、沙特阿拉伯、墨西哥、波兰
D组：法国、澳大利亚、丹麦、突尼斯
E组：西班牙、哥斯达黎加、德国、日本
F组：比利时、加拿大、摩洛哥、克罗地亚
G组：巴西、塞尔维亚、瑞士、喀麦隆
H组：葡萄牙、加纳、乌拉圭、韩国

话不多说，开始我们的预测~

冠军球队预测

采用的数据集：
1872年至2022年国际足球成绩：

shootouts.csv
results.csv

国际足联世界排名1992-2022

fifa_ranking-2022-10-06.csv

2022年国际足联世界杯分组

Fifa_Worldcup_2022_Groups.csv

引入所需包：

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
import plotly.graph_objs as go
import plotly.figure_factory as ff
import plotly.express as px
from datetime import datetime
import warnings
import plotly.graph_objs as go导入历史对阵成绩数据：

df = pd.read_csv(&#34;/home/results.csv&#34;)
df

导入历史排名数据，并对rank_date做变换和筛选：

rank = pd.read_csv(&#39;/home/fifa_ranking-2022-10-06.csv&#39;)

rank[&#34;rank_date&#34;] = pd.to_datetime(rank[&#34;rank_date&#34;])
rank = rank[(rank[&#34;rank_date&#34;] >= &#34;2020-1-1&#34;)].reset_index(drop=True)

简单的数据处理和清洗：

rank[&#34;country_full&#34;] = rank[&#34;country_full&#34;].str.replace(&#34;IR Iran&#34;, &#34;Iran&#34;).str.replace(&#34;Korea Republic&#34;, &#34;South Korea&#34;).str.replace(&#34;USA&#34;, &#34;United States&#34;)
rank = rank.set_index([&#39;rank_date&#39;]).groupby([&#39;country_full&#39;], group_keys=False).resample(&#39;D&#39;).first().fillna(method=&#39;ffill&#39;).reset_index()

df[&#34;date&#34;] = pd.to_datetime(df[&#34;date&#34;])
df = df[(df[&#34;date&#34;] >= &#34;2020-1-1&#34;)].reset_index(drop=True)
df.head()

数据转换透视表：

ranking_df = pd.pivot_table(data = rank,
                        values = &#39;total_points&#39;,
                        index = &#39;country_full&#39;,
                        columns = &#39;rank_date&#39;).fillna(0.0)
ranking_df.head()

当今10支最佳球队的国际足联排名的演变：

best_ranks = ranking_df.loc[ranking_df[&#39;2022-10-06&#39;].sort_values(ascending = False)[:10].index]
fig = go.Figure()

for i in range(len(best_ranks.values)):
fig.add_trace(go.Scatter(x = best_ranks.columns,
                           y = best_ranks.iloc,
                           name = best_ranks.index))

fig.update_layout(
title=&#34;Evolution of the FIFA Ranking for today\&#39;s 10 best teams&#34;,
yaxis_title=&#34;Points&#34;
)
fig.show()

十大足球比赛的类型：

tournament = df[&#39;tournament&#39;].value_counts()
tournament = tournament[:10]

plt.figure(figsize = (15,10))
sns.set_style(&#34;whitegrid&#34;)
ax = sns.barplot(y=tournament.index, x=tournament.values, palette=&#34;Reds_r&#34;, orient=&#39;h&#39;)
ax.set_ylabel(&#39;Tournament&#39;, size=16)
ax.set_xlabel(&#39;Number of tournament&#39;, size=16)
ax.set_title(&#34;TOP 10 TYPE OF MATCH TOURNAMENTS&#34;, fontsize=18)

场均进球数前10名的球队：

home = df[[&#39;home_team&#39;, &#39;home_score&#39;]].rename(columns={&#39;home_team&#39;:&#39;team&#39;, &#39;home_score&#39;:&#39;score&#39;})
away = df[[&#39;away_team&#39;, &#39;away_score&#39;]].rename(columns={&#39;away_team&#39;:&#39;team&#39;, &#39;away_score&#39;:&#39;score&#39;})

team_score = home.append(away).reset_index(drop=True)

country_info = team_score.groupby(&#39;team&#39;)[&#39;score&#39;].agg([&#39;sum&#39;,&#39;count&#39;,&#39;mean&#39;]).reset_index()
country_info = country_info.rename(columns={&#39;sum&#39;:&#39;nb_goals&#39;, &#39;count&#39;:&#39;nb_matches&#39;, &#39;mean&#39;:&#39;goal_avg&#39;})
plt.figure(figsize = (15,10))
sns.set_style(&#34;whitegrid&#34;)
plt_data = country_info.sort_values(by=&#39;goal_avg&#39;, ascending=False)[:10]
ax = sns.barplot(x=&#34;team&#34;, y=&#34;goal_avg&#34;, data=plt_data, palette=&#34;Blues_r&#34;)
ax.set_xlabel(&#39;Team&#39;, size=16)
ax.set_ylabel(&#39;Goal average per match&#39;, size=16)
ax.set_title(&#34;TOP 10 OF GOAL AVERAGE PER MATCH&#34;, fontsize=18)

进球数前10名的球队：

plt.figure(figsize = (15,10))
sns.set_style(&#34;whitegrid&#34;)
plt_data = country_info.sort_values(by=&#39;nb_goals&#39;, ascending=False)[:10]
ax = sns.barplot(x=&#34;team&#34;, y=&#34;nb_goals&#34;, data=plt_data, palette=&#34;Blues_r&#34;)
ax.set_xlabel(&#39;Team&#39;, size=16)
ax.set_ylabel(&#39;Number of goals&#39;, size=16)
ax.set_title(&#34;TOP 10 OF NUMBER OF GOALS&#34;, fontsize=18)

del plt_data, country_info

导入卡塔尔世界杯的分组对阵数据：

world_cup = pd.read_csv(&#34;/home/Fifa_Worldcup_2022_Groups.csv&#34;)
world_cup

简单的数据处理和清洗：

world_cup = world_cup.replace({&#34;Korea Republic&#34; : &#34;South Korea&#34;,&#34;USA&#34;: &#34;United States&#34;})
world_cup = world_cup.set_index(&#39;Team&#39;)与rank表进行关联：

home_team：主队

away_team：客队

df = df.merge(rank,left_on=[&#39;date&#39;, &#39;home_team&#39;],right_on=[&#39;rank_date&#39;, &#39;country_full&#39;])
df = df.merge(rank, left_on=[&#39;date&#39;, &#39;away_team&#39;], right_on=[&#39;rank_date&#39;, &#39;country_full&#39;], suffixes=(&#39;_home&#39;, &#39;_away&#39;)) 查看df所有特征变量的相关性：

fig, ax = plt.subplots()
fig.set_size_inches(20, 20)
corr1 = df.corr()
corr1
sns.heatmap(corr1,annot=True)

数据变换和特征处理：

df[&#39;rank_difference&#39;] = df[&#39;rank_home&#39;] - df[&#39;rank_away&#39;]
df[&#39;average_rank&#39;] = (df[&#39;rank_home&#39;] + df[&#39;rank_away&#39;])/2
df[&#39;point_difference&#39;] = df[&#39;total_points_home&#39;] - df[&#39;total_points_away&#39;]
df[&#39;score_difference&#39;] = df[&#39;home_score&#39;] - df[&#39;away_score&#39;]
df[&#39;is_won&#39;] = df[&#39;score_difference&#39;] > 0
df[&#39;is_stake&#39;] = df[&#39;tournament&#39;] != &#39;Friendly&#39;

df[&#39;wc_participant&#39;] = df[&#39;home_team&#39;] * df[&#39;home_team&#39;].isin(world_cup.index.tolist())
df[&#39;wc_participant&#39;] = df[&#39;wc_participant&#39;].replace({&#39;&#39;:&#39;Other&#39;})
df = df.join(pd.get_dummies(df[&#39;wc_participant&#39;]))导入机器学习所需的包：

from sklearn import preprocessing
from sklearn.model_selection import train_test_split
from scipy import stats
from sklearn.linear_model import LogisticRegression
from collections import Counter
from sklearn.metrics import confusion_matrix
from sklearn.metrics import accuracy_score, f1_score
from xgboost import XGBClassifier
from sklearn.ensemble import RandomForestRegressor
from sklearn.naive_bayes import BernoulliNB
from sklearn.naive_bayes import GaussianNB
from sklearn.svm import SVC
from sklearn.ensemble import RandomForestRegressor
import warnings拆分数据集：

X, y = df.loc[:,[&#39;average_rank&#39;, &#39;rank_difference&#39;, &#39;point_difference&#39;, &#39;is_stake&#39;]], df[&#39;is_won&#39;]
X_train, X_test, y_train, y_test = train_test_split(
X, y, test_size=0.2, random_state=42)定义一个画ROC曲线的函数plot_roc_cur：

def plot_roc_cur(fper, tper):
plt.plot(fper, tper, color=&#39;orange&#39;, label=&#39;ROC&#39;)
plt.plot([0, 1], [0, 1], color=&#39;darkblue&#39;, linestyle=&#39;--&#39;)
plt.xlabel(&#39;False Positive Rate&#39;)
plt.ylabel(&#39;True Positive Rate&#39;)
plt.title(&#39;Receiver Operating Characteristic (ROC) Curve&#39;)
plt.legend()
plt.show()定义一个支持跑多个模型的函数run_model：
通过run_model可以训练模型，并且打印和输出模型的效果评估指标，比如模型的准确率、混淆矩阵、ROC曲线以及时间消耗等等。（关于模型评估指标可以参考小编之前的文章《聊聊机器学习分类问题常用的评价指标》）

import time
from sklearn.metrics import accuracy_score, roc_auc_score, cohen_kappa_score, plot_confusion_matrix, roc_curve, classification_report
def run_model(model, X_train, y_train, X_test, y_test, verbose=True):
t0=time.time()
if verbose == False:
      model.fit(X_train,y_train, verbose=0)
else:
      model.fit(X_train,y_train)
y_pred = model.predict(X_test)
accuracy = accuracy_score(y_test, y_pred)
roc_auc = roc_auc_score(y_test, y_pred)
coh_kap = cohen_kappa_score(y_test, y_pred)
time_taken = time.time()-t0
print(&#34;Accuracy = {}&#34;.format(accuracy))
print(&#34;ROC Area under Curve = {}&#34;.format(roc_auc))
print(&#34;Cohen&#39;s Kappa = {}&#34;.format(coh_kap))
print(&#34;Time taken = {}&#34;.format(time_taken))
print(classification_report(y_test,y_pred,digits=5))

probs = model.predict_proba(X_test)
probs = probs[:, 1]
fper, tper, thresholds = roc_curve(y_test, probs)
plot_roc_cur(fper, tper)

plot_confusion_matrix(model, X_test, y_test,cmap=plt.cm.Blues, normalize = &#39;all&#39;)

return model, accuracy, roc_auc, coh_kap, time_taken逻辑回归的效果评估：

from sklearn.linear_model import LogisticRegression

params_lr = {&#39;penalty&#39;: &#39;l1&#39;, &#39;solver&#39;:&#39;liblinear&#39;}

model_lr = LogisticRegression(**params_lr)
model_lr, accuracy_lr, roc_auc_lr, coh_kap_lr, tt_lr = run_model(model_lr, X_train, y_train, X_test, y_test)

决策树的效果评估：

from sklearn.tree import DecisionTreeClassifier

params_dt = {&#39;max_depth&#39;: 16,
         &#39;max_features&#39;: &#34;sqrt&#34;}

model_dt = DecisionTreeClassifier(**params_dt)
model_dt, accuracy_dt, roc_auc_dt, coh_kap_dt, tt_dt = run_model(model_dt, X_train, y_train, X_test, y_test)

神经网络MLPClassifier的效果评估：

from sklearn.neural_network import MLPClassifier

params_nn = {&#39;hidden_layer_sizes&#39;: (30,30,30),
         &#39;activation&#39;: &#39;logistic&#39;,
         &#39;solver&#39;: &#39;lbfgs&#39;,
         &#39;max_iter&#39;: 500}

model_nn = MLPClassifier(**params_nn)
model_nn, accuracy_nn, roc_auc_nn, coh_kap_nn, tt_nn = run_model(model_nn, X_train, y_train, X_test, y_test)

随机森林的效果评估：

from sklearn.ensemble import RandomForestClassifier

params_rf = {&#39;max_depth&#39;: 16,
         &#39;min_samples_leaf&#39;: 1,
         &#39;min_samples_split&#39;: 2,
         &#39;n_estimators&#39;: 100,
         &#39;random_state&#39;: 12345}

model_rf = RandomForestClassifier(**params_rf)
model_rf, accuracy_rf, roc_auc_rf, coh_kap_rf, tt_rf = run_model(model_rf, X_train, y_train, X_test, y_test)

LightGBM的效果评估：

import lightgbm as lgb
params_lgb ={&#39;colsample_bytree&#39;: 0.95,
      &#39;max_depth&#39;: 16,
      &#39;min_split_gain&#39;: 0.1,
      &#39;n_estimators&#39;: 200,
      &#39;num_leaves&#39;: 50,
      &#39;reg_alpha&#39;: 1.2,
      &#39;reg_lambda&#39;: 1.2,
      &#39;subsample&#39;: 0.95,
      &#39;subsample_freq&#39;: 20}

model_lgb = lgb.LGBMClassifier(**params_lgb)
model_lgb, accuracy_lgb, roc_auc_lgb, coh_kap_lgb, tt_lgb = run_model(model_lgb, X_train, y_train, X_test, y_test)

xgboost的效果评估：

import xgboost as xgb
params_xgb ={&#39;n_estimators&#39;: 1000,
         &#39;max_depth&#39;: 128}

model_xgb = xgb.XGBClassifier(**params_xgb)
model_xgb, accuracy_xgb, roc_auc_xgb, coh_kap_xgb, tt_xgb = run_model(model_xgb, X_train, y_train, X_test, y_test)

各支球队世界杯时的排名：

world_cup_rankings = rank.loc[(rank[&#39;rank_date&#39;] == rank[&#39;rank_date&#39;].max()) &
                                 rank[&#39;country_full&#39;].isin(world_cup.index.unique())]
world_cup_rankings = world_cup_rankings.set_index([&#39;country_full&#39;])小组赛预测：

from itertools import combinations

opponents = [&#39;First match \nagainst&#39;, &#39;Second match\n against&#39;, &#39;Third match\n against&#39;]

world_cup[&#39;points&#39;] = 0
world_cup[&#39;total_prob&#39;] = 0

for group in set(world_cup[&#39;Group&#39;]):
print(&#39;___Starting group {}:___&#39;.format(group))
for home, away in combinations(world_cup.query(&#39;Group == &#34;{}&#34;&#39;.format(group)).index, 2):
      print(&#34;{} vs. {}: &#34;.format(home, away), end=&#39;&#39;)
      row = pd.DataFrame(np.array([[np.nan, np.nan, np.nan, True]]), columns=X_test.columns)
      home_rank = world_cup_rankings.loc[home, &#39;rank&#39;]
      home_points = world_cup_rankings.loc[home, &#39;total_points&#39;]
      opp_rank = world_cup_rankings.loc[away, &#39;rank&#39;]
      opp_points = world_cup_rankings.loc[away, &#39;total_points&#39;]
      row[&#39;average_rank&#39;] = (home_rank + opp_rank) / 2
      row[&#39;rank_difference&#39;] = home_rank - opp_rank
      row[&#39;point_difference&#39;] = home_points - opp_points

      home_win_prob = model_lr.predict_proba(row)[:,1][0]
      world_cup.loc[home, &#39;total_prob&#39;] += home_win_prob
      world_cup.loc[away, &#39;total_prob&#39;] += 1-home_win_prob

      points = 0
      if home_win_prob <= 0.5 - margin:
         print(&#34;{} wins with {:.2f}&#34;.format(away, 1-home_win_prob))
         world_cup.loc[away, &#39;points&#39;] += 3
      if home_win_prob > 0.5 - margin:
         points = 1
      if home_win_prob >= 0.5 + margin:
         points = 3
         world_cup.loc[home, &#39;points&#39;] += 3
         print(&#34;{} wins with {:.2f}&#34;.format(home, home_win_prob))
      if points == 1:
         print(&#34;Draw&#34;)
         world_cup.loc[home, &#39;points&#39;] += 1
         world_cup.loc[away, &#39;points&#39;] += 1___Starting group F:___
Belgium vs. Canada: Belgium wins with 0.85
Belgium vs. Morocco: Belgium wins with 0.77
Belgium vs. Croatia: Belgium wins with 0.67
Canada vs. Morocco: Morocco wins with 0.69
Canada vs. Croatia: Croatia wins with 0.78
Morocco vs. Croatia: Croatia wins with 0.69
___Starting group C:___
Argentina vs. Saudi Arabia: Argentina wins with 0.84
Argentina vs. Mexico: Argentina wins with 0.61
Argentina vs. Poland: Argentina wins with 0.74
Saudi Arabia vs. Mexico: Mexico wins with 0.81
Saudi Arabia vs. Poland: Poland wins with 0.71
Mexico vs. Poland: Mexico wins with 0.56
___Starting group A:___
Qatar vs. Ecuador: Ecuador wins with 0.61
Qatar vs. Senegal: Senegal wins with 0.75
Qatar vs. Netherlands: Netherlands wins with 0.85
Ecuador vs. Senegal: Senegal wins with 0.73
Ecuador vs. Netherlands: Netherlands wins with 0.84
Senegal vs. Netherlands: Netherlands wins with 0.73
___Starting group D:___
France vs. Denmark: France wins with 0.56
France vs. Tunisia: France wins with 0.77
France vs. Australia: France wins with 0.78
Denmark vs. Tunisia: Denmark wins with 0.65
Denmark vs. Australia: Denmark wins with 0.67
Tunisia vs. Australia: Australia wins with 0.55
___Starting group H:___
Portugal vs. Ghana: Portugal wins with 0.79
Portugal vs. Uruguay: Draw
Portugal vs. South Korea: Portugal wins with 0.63
Ghana vs. Uruguay: Uruguay wins with 0.84
Ghana vs. South Korea: South Korea wins with 0.74
Uruguay vs. South Korea: Uruguay wins with 0.58
___Starting group E:___
Spain vs. Germany: Draw
Spain vs. Japan: Spain wins with 0.65
Spain vs. Costa Rica: Spain wins with 0.72
Germany vs. Japan: Germany wins with 0.55
Germany vs. Costa Rica: Germany wins with 0.63
Japan vs. Costa Rica:
Draw
___Starting group B:___
England vs. Iran: England wins with 0.66
England vs. United States: England wins with 0.57
England vs. Wales: England wins with 0.65
Iran vs. United States: United States wins with 0.67
Iran vs. Wales: Wales wins with 0.58
United States vs. Wales: Draw
___Starting group G:___
Brazil vs. Serbia: Brazil wins with 0.80
Brazil vs. Switzerland: Brazil wins with 0.72
Brazil vs. Cameroon: Brazil wins with 0.87
Serbia vs. Switzerland: Switzerland wins with 0.68
Serbia vs. Cameroon: Draw
Switzerland vs. Cameroon: Switzerland wins with 0.65

淘汰赛和总决赛预测：

pairing = [0,3,4,7,8,11,12,15,1,2,5,6,9,10,13,14]

world_cup = world_cup.sort_values(by=[&#39;Group&#39;, &#39;points&#39;, &#39;total_prob&#39;], ascending=False).reset_index()
next_round_wc = world_cup.groupby(&#39;Group&#39;).nth([0, 1]) # select the top 2
next_round_wc = next_round_wc.reset_index()
next_round_wc = next_round_wc.loc[pairing]
next_round_wc = next_round_wc.set_index(&#39;Team&#39;)

finals = [&#39;round_of_16&#39;, &#39;quarterfinal&#39;, &#39;semifinal&#39;, &#39;final&#39;]

labels = list()
odds = list()

for f in finals:
print(&#34;___Starting of the {}___&#34;.format(f))
iterations = int(len(next_round_wc) / 2)
winners = []

for i in range(iterations):
      home = next_round_wc.index[i*2]
      away = next_round_wc.index[i*2+1]
      print(&#34;{} vs. {}: &#34;.format(home,
                                 away),
                                 end=&#39;&#39;)
      row = pd.DataFrame(np.array([[np.nan, np.nan, np.nan, True]]), columns=X_test.columns)
      home_rank = world_cup_rankings.loc[home, &#39;rank&#39;]
      home_points = world_cup_rankings.loc[home, &#39;total_points&#39;]
      opp_rank = world_cup_rankings.loc[away, &#39;rank&#39;]
      opp_points = world_cup_rankings.loc[away, &#39;total_points&#39;]
      row[&#39;average_rank&#39;] = (home_rank + opp_rank) / 2
      row[&#39;rank_difference&#39;] = home_rank - opp_rank
      row[&#39;point_difference&#39;] = home_points - opp_points

      home_win_prob = model_lr.predict_proba(row)[:,1][0]
      if model_lr.predict_proba(row)[:,1] <= 0.5:
         print(&#34;{0} wins with probability {1:.2f}&#34;.format(away, 1-home_win_prob))
         winners.append(away)
      else:
         print(&#34;{0} wins with probability {1:.2f}&#34;.format(home, home_win_prob))
         winners.append(home)

      labels.append(&#34;{}({:.2f}) vs. {}({:.2f})&#34;.format(world_cup_rankings.loc[home, &#39;country_abrv&#39;],
                                                      1/home_win_prob,
                                                      world_cup_rankings.loc[away, &#39;country_abrv&#39;],
                                                      1/(1-home_win_prob)))
      odds.append([home_win_prob, 1-home_win_prob])

next_round_wc = next_round_wc.loc[winners]
print(&#34;\n&#34;)___Starting of the round_of_16___
Netherlands vs. United States: Netherlands wins with probability 0.52
Argentina vs. Denmark: Argentina wins with probability 0.58
Spain vs. Croatia: Spain wins with probability 0.53
Brazil vs. Portugal: Brazil wins with probability 0.67
Senegal vs. England: England wins with probability 0.77
Mexico vs. France: France wins with probability 0.73
Germany vs. Belgium: Belgium wins with probability 0.79
Switzerland vs. Uruguay: Uruguay wins with probability 0.58
___Starting of the quarterfinal___
Netherlands vs. Argentina: Argentina wins with probability 0.69
Spain vs. Brazil: Brazil wins with probability 0.75
England vs. France: France wins with probability 0.63
Belgium vs. Uruguay: Belgium wins with probability 0.68
___Starting of the semifinal___
Argentina vs. Brazil: Brazil wins with probability 0.68
France vs. Belgium: Belgium wins with probability 0.66
___Starting of the final___
Brazil vs. Belgium: Belgium wins with probability 0.54

我们看到模型预测的结果，决赛将由夺冠热门巴西队对阵拥有卢卡库、阿扎尔、德布劳内等多名明星球员的比利时队，最终比利时队赢得冠军！（预测结果仅供参考，一切结果以实际为准）

最后我们贴上用SmartNoteBook创作的一幅淘汰赛晋级图：

《新语数据故事汇，数说新语》科普数据科学、讲述数据故事，深层次挖掘数据价值。欢迎各位朋友投稿！

微信号｜SnbDataStory
《新语数据故事汇，数说新语》

本文使用文章同步助手同步

		自动登录	找回密码
密码			立即注册