深度学习测试

import pandas as pd
import  numpy as np
import tflearn
from sklearn.preprocessing import LabelEncoder,OneHotEncoder
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.tree import DecisionTreeClassifier as DTC
from sklearn.metrics import classification_report
from sklearn.ensemble import RandomForestClassifier as RFC
#数据预处理
data=pd.read_csv(r'D:\Python\customer_churn.csv')
df=data.iloc[:,3:]
x=df.iloc[:,:10]
y=df.Exited
x.Geography=LabelEncoder().fit_transform(x.Geography)#标签化
x.Gender=LabelEncoder().fit_transform(x.Gender)
x=OneHotEncoder(categorical_features=[1]).fit_transform(x).toarray()#独热码
# y= y[:, np.newaxis]
y=y.values.reshape(10000,1)
y=OneHotEncoder().fit_transform(y).toarray()
x_train,x_test,y_train,y_test=train_test_split(x,y,test_size=0.2,random_state=1)#random_state表示随机种子
x_train=StandardScaler().fit_transform(x_train)
x_test=StandardScaler().fit_transform(x_test)

# clf=DTC().fit(x_train,y_train)#决策树
clf=RFC(n_estimators=50).fit(x_train,y_train)#随机森林
result=classification_report(y_test,clf.predict(x_test))
# print(result)
net=tflearn.input_data(shape=[None,11])#输入层
net=tflearn.fully_connected(net,6,activation='relu')#隐藏层1
net=tflearn.fully_connected(net,6,activation='relu')#隐藏层2
net=tflearn.fully_connected(net,6,activation='relu')#隐藏层3
net = tflearn.fully_connected(net, n_units=2, activation='softmax')#输出层2表示输出单元
net = tflearn.regression(net)
model=tflearn.DNN(net)
model.fit(x_train, y_train, n_epoch=30, batch_size=32, show_metric=True)#n_epoch：数据训练几个轮次batch_size：每一次输入给模型的数据行数show_metric：训练过程中要不要打印结果