# 通用的預處理框架import pandas as pd
import numpy as np
import scipy as sp
# 文件讀取defread_csv_file(f, logging=False):print("==========讀取數據=========")
data = pd.read_csv(f)if logging:print(data.head(5))print(f,"包含以下列")print(data.columns.values)print(data.describe())print(data.info())return data
LR(線性迴歸)
# 通用的LogisticRegression框架import pandas as pd
import numpy as np
from scipy import sparse
from sklearn.preprocessing import OneHotEncoder
from sklearn.linear_model import LogisticRegression
from sklearn.preprocessing import StandardScaler
# 1. load data
df_train = pd.DataFrame()
df_test = pd.DataFrame()
y_train = df_train['label'].values
# 2. process data
ss = StandardScaler()# 3. feature engineering/encoding# 3.1 For Labeled Feature
enc = OneHotEncoder()
feats =["creativeID","adID","campaignID"]for i, feat inenumerate(feats):
x_train = enc.fit_transform(df_train[feat].values.reshape(-1,1))
x_test = enc.fit_transform(df_test[feat].values.reshape(-1,1))if i ==0:
X_train, X_test = x_train, x_test
else:
X_train, X_test = sparse.hstack((X_train, x_train)), sparse.hstack((X_test, x_test))# 3.2 For Numerical Feature# It must be a 2-D Data for StandardScalar, otherwise reshape(-1, len(feats)) is required
feats =["price","age"]
x_train = ss.fit_transform(df_train[feats].values)
x_test = ss.fit_transform(df_test[feats].values)
X_train, X_test = sparse.hstack((X_train, x_train)), sparse.hstack((X_test, x_test))# model training
lr = LogisticRegression()
lr.fit(X_train, y_train)
proba_test = lr.predict_proba(X_test)[:,1]
LightBGM
二分類
import lightgbm as lgb
import pandas as pd
import numpy as np
import pickle
from sklearn.metrics import roc_auc_score
from sklearn.model_selection import train_test_split
print("Loading Data ... ")# 導入數據
train_x, train_y, test_x = load_data()# 用sklearn.cross_validation進行訓練數據集劃分,這裏訓練集和交叉驗證集比例爲7:3,可以自己根據需要設置
X, val_X, y, val_y = train_test_split(
train_x,
train_y,
test_size=0.05,
random_state=1,
stratify=train_y ## 這裏保證分割後y的比例分佈與原數據一致)
X_train = X
y_train = y
X_test = val_X
y_test = val_y
# create dataset for lightgbm
lgb_train = lgb.Dataset(X_train, y_train)
lgb_eval = lgb.Dataset(X_test, y_test, reference=lgb_train)# specify your configurations as a dict
params ={'boosting_type':'gbdt','objective':'binary','metric':{'binary_logloss','auc'},'num_leaves':5,'max_depth':6,'min_data_in_leaf':450,'learning_rate':0.1,'feature_fraction':0.9,'bagging_fraction':0.95,'bagging_freq':5,'lambda_l1':1,'lambda_l2':0.001,# 越小l2正則程度越高'min_gain_to_split':0.2,'verbose':5,'is_unbalance':True}# trainprint('Start training...')
gbm = lgb.train(params,
lgb_train,
num_boost_round=10000,
valid_sets=lgb_eval,
early_stopping_rounds=500)print('Start predicting...')
preds = gbm.predict(test_x, num_iteration=gbm.best_iteration)# 輸出的是概率結果# 導出結果
threshold =0.5for pred in preds:
result =1if pred > threshold else0# 導出特徵重要性
importance = gbm.feature_importance()
names = gbm.feature_name()withopen('./feature_importance.txt','w+')asfile:for index, im inenumerate(importance):
string = names[index]+', '+str(im)+'\n'file.write(string)
多分類
import lightgbm as lgb
import pandas as pd
import numpy as np
import pickle
from sklearn.metrics import roc_auc_score
from sklearn.model_selection import train_test_split
print("Loading Data ... ")# 導入數據
train_x, train_y, test_x = load_data()# 用sklearn.cross_validation進行訓練數據集劃分,這裏訓練集和交叉驗證集比例爲7:3,可以自己根據需要設置
X, val_X, y, val_y = train_test_split(
train_x,
train_y,
test_size=0.05,
random_state=1,
stratify=train_y ## 這裏保證分割後y的比例分佈與原數據一致)
X_train = X
y_train = y
X_test = val_X
y_test = val_y
# create dataset for lightgbm
lgb_train = lgb.Dataset(X_train, y_train)
lgb_eval = lgb.Dataset(X_test, y_test, reference=lgb_train)# specify your configurations as a dict
params ={'boosting_type':'gbdt','objective':'multiclass','num_class':9,'metric':'multi_error','num_leaves':300,'min_data_in_leaf':100,'learning_rate':0.01,'feature_fraction':0.8,'bagging_fraction':0.8,'bagging_freq':5,'lambda_l1':0.4,'lambda_l2':0.5,'min_gain_to_split':0.2,'verbose':5,'is_unbalance':True}# trainprint('Start training...')
gbm = lgb.train(params,
lgb_train,
num_boost_round=10000,
valid_sets=lgb_eval,
early_stopping_rounds=500)print('Start predicting...')
preds = gbm.predict(test_x, num_iteration=gbm.best_iteration)# 輸出的是概率結果# 導出結果for pred in preds:
result = prediction =int(np.argmax(pred))# 導出特徵重要性
importance = gbm.feature_importance()
names = gbm.feature_name()withopen('./feature_importance.txt','w+')asfile:for index, im inenumerate(importance):
string = names[index]+', '+str(im)+'\n'file.write(string)
XGB
二分類
import numpy as np
import pandas as pd
import xgboost as xgb
import time
from sklearn.model_selection import StratifiedKFold
from sklearn.model_selection import train_test_split
train_x, train_y, test_x = load_data()# 構建特徵# 用sklearn.cross_validation進行訓練數據集劃分,這裏訓練集和交叉驗證集比例爲7:3,可以自己根據需要設置
X, val_X, y, val_y = train_test_split(
train_x,
train_y,
test_size=0.01,
random_state=1,
stratify=train_y
)# xgb矩陣賦值
xgb_val = xgb.DMatrix(val_X, label=val_y)
xgb_train = xgb.DMatrix(X, label=y)
xgb_test = xgb.DMatrix(test_x)# xgboost模型 #####################
params ={'booster':'gbtree',# 'objective': 'multi:softmax', # 多分類的問題、# 'objective': 'multi:softprob', # 多分類概率'objective':'binary:logistic','eval_metric':'logloss',# 'num_class': 9, # 類別數,與 multisoftmax 並用'gamma':0.1,# 用於控制是否後剪枝的參數,越大越保守,一般0.1、0.2這樣子。'max_depth':8,# 構建樹的深度,越大越容易過擬合'alpha':0,# L1正則化係數'lambda':10,# 控制模型複雜度的權重值的L2正則化項參數,參數越大,模型越不容易過擬合。'subsample':0.7,# 隨機採樣訓練樣本'colsample_bytree':0.5,# 生成樹時進行的列採樣'min_child_weight':3,# 這個參數默認是 1,是每個葉子裏面 h 的和至少是多少,對正負樣本不均衡時的 0-1 分類而言# ,假設 h 在 0.01 附近,min_child_weight 爲 1 意味着葉子節點中最少需要包含 100 個樣本。# 這個參數非常影響結果,控制葉子節點中二階導的和的最小值,該參數值越小,越容易 overfitting。'silent':0,# 設置成1則沒有運行信息輸出,最好是設置爲0.'eta':0.03,# 如同學習率'seed':1000,'nthread':-1,# cpu 線程數'missing':1,'scale_pos_weight':(np.sum(y==0)/np.sum(y==1))# 用來處理正負樣本不均衡的問題,通常取:sum(negative cases) / sum(positive cases)# 'eval_metric': 'auc'}
plst =list(params.items())
num_rounds =2000# 迭代次數
watchlist =[(xgb_train,'train'),(xgb_val,'val')]# 交叉驗證
result = xgb.cv(plst, xgb_train, num_boost_round=200, nfold=4, early_stopping_rounds=200, verbose_eval=True, folds=StratifiedKFold(n_splits=4).split(X, y))# 訓練模型並保存# early_stopping_rounds 當設置的迭代次數較大時,early_stopping_rounds 可在一定的迭代次數內準確率沒有提升就停止訓練
model = xgb.train(plst, xgb_train, num_rounds, watchlist, early_stopping_rounds=200)
model.save_model('../data/model/xgb.model')# 用於存儲訓練出的模型
preds = model.predict(xgb_test)# 導出結果
threshold =0.5for pred in preds:
result =1if pred > threshold else0
Keras
二分類
import numpy as np
import pandas as pd
import time
from sklearn.model_selection import train_test_split
from matplotlib import pyplot as plt
from keras.models import Sequential
from keras.layers import Dropout
from keras.layers import Dense, Activation
from keras.utils.np_utils import to_categorical
# coding=utf-8from model.util import load_data as load_data_1
from model.util_combine_train_test import load_data as load_data_2
from sklearn.preprocessing import StandardScaler # 用於特徵的標準化from sklearn.preprocessing import Imputer
print("Loading Data ... ")# 導入數據
train_x, train_y, test_x = load_data()# 構建特徵
X_train = train_x.values
X_test = test_x.values
y = train_y
imp = Imputer(missing_values='NaN', strategy='mean', axis=0)
X_train = imp.fit_transform(X_train)
sc = StandardScaler()
sc.fit(X_train)
X_train = sc.transform(X_train)
X_test = sc.transform(X_test)
model = Sequential()
model.add(Dense(256, input_shape=(X_train.shape[1],)))
model.add(Activation('tanh'))
model.add(Dropout(0.3))
model.add(Dense(512))
model.add(Activation('relu'))
model.add(Dropout(0.3))
model.add(Dense(512))
model.add(Activation('tanh'))
model.add(Dropout(0.3))
model.add(Dense(256))
model.add(Activation('linear'))
model.add(Dense(1))# 這裏需要和輸出的維度一致
model.add(Activation('sigmoid'))# For a multi-class classification problem
model.compile(loss='binary_crossentropy',
optimizer='rmsprop',
metrics=['accuracy'])
epochs =100
model.fit(X_train, y, epochs=epochs, batch_size=2000, validation_split=0.1, shuffle=True)# 導出結果
threshold =0.5for index, case inenumerate(X_test):
case =np.array([case])
prediction_prob = model.predict(case)
prediction =1if prediction_prob[0][0]> threshold else0
多分類
import numpy as np
import pandas as pd
import time
from sklearn.model_selection import train_test_split
from matplotlib import pyplot as plt
from keras.models import Sequential
from keras.layers import Dropout
from keras.layers import Dense, Activation
from keras.utils.np_utils import to_categorical
# coding=utf-8from model.util import load_data as load_data_1
from model.util_combine_train_test import load_data as load_data_2
from sklearn.preprocessing import StandardScaler # 用於特徵的標準化from sklearn.preprocessing import Imputer
print("Loading Data ... ")# 導入數據
train_x, train_y, test_x = load_data()# 構建特徵
X_train = train_x.values
X_test = test_x.values
y = train_y
# 特徵處理
sc = StandardScaler()
sc.fit(X_train)
X_train = sc.transform(X_train)
X_test = sc.transform(X_test)
y = to_categorical(y)## 這一步很重要,一定要將多類別的標籤進行one-hot編碼
model = Sequential()
model.add(Dense(256, input_shape=(X_train.shape[1],)))
model.add(Activation('tanh'))
model.add(Dropout(0.3))
model.add(Dense(512))
model.add(Activation('relu'))
model.add(Dropout(0.3))
model.add(Dense(512))
model.add(Activation('tanh'))
model.add(Dropout(0.3))
model.add(Dense(256))
model.add(Activation('linear'))
model.add(Dense(9))# 這裏需要和輸出的維度一致
model.add(Activation('softmax'))# For a multi-class classification problem
model.compile(optimizer='rmsprop',
loss='categorical_crossentropy',
metrics=['accuracy'])
epochs =200
model.fit(X_train, y, epochs=epochs, batch_size=200, validation_split=0.1, shuffle=True)# 導出結果for index, case inenumerate(X_test):
case = np.array([case])
prediction_prob = model.predict(case)
prediction = np.argmax(prediction_prob)