python bug分析

ValueError: Classification metrics can‘t handle a mix of binary and continuous targets

问题分析

源代码

import numpy as np
import matplotlib.pyplot as plt
#%%
import pandas as pd
from sklearn.preprocessing import LabelEncoder,OneHotEncoder
from sklearn.compose import ColumnTransformer

dataset = pd.read_csv("F:\数模\美赛\Machine Learning A-Z Chinese Template Folder\Part 8 - Deep Learning\Section 32 - Artificial Neural Networks (ANN)\Churn_Modelling.csv")
x=dataset.iloc[:,3:13].values
y=dataset.iloc[:,13].values
#处理虚拟变量
labelencoder_x= LabelEncoder();
x[:,1]= labelencoder_x.fit_transform(x[:,1])
x[:,2]=labelencoder_x.fit_transform(x[:,2])
ct = ColumnTransformer([('one_hot_encoder', OneHotEncoder(), [1])], remainder='passthrough')
x = ct.fit_transform(x)
#注意虚拟变量陷阱
x=x[:,1:]

from sklearn.model_selection import train_test_split

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=0)

# 数据的特征缩放
from sklearn.preprocessing import StandardScaler

sc_x = StandardScaler()

x_train = sc_x.fit_transform(x_train)
x_test = sc_x.transform(x_test)

import keras
from keras.models import Sequential
from keras.layers import Dense
#初始化ann
classifier=Sequential()
#添加输入层和第一层隐藏层
#x的列数就是输入神经元的个数
#一般隐藏层的神经元个数是输入层和输出层的平均数
classifier.add(Dense(units=6,kernel_initializer='uniform',activation='relu',input_dim=11))
#添加第二层隐藏层
classifier.add(Dense(units=6,kernel_initializer='uniform',activation='relu'))
#添加输出层,有三个或以上分类结果时改动units输出层个数和activation改为softmax
classifier.add(Dense(units=1,kernel_initializer='uniform',activation='sigmoid'))
#编译ann,不同输出层的激活函数有不同的损失函数,当分类结果大于等于三个时损失函数用categorical
classifier.compile(optimizer='adam',loss='binary_crossentropy',metrics=['accuracy'])
classifier.fit(x_train,y_train,batch_size=5,epochs=100)
#%%
#运用神经网络预测测试集的结果
y_pred=classifier.predict(x_test)
y_pred=(y_pred>0.5)
#利用混淆矩阵判断预测准确性
from sklearn.metrics import confusion_matrix
cm=confusion_matrix(y_test,y_pred)

分析报错原因是用python console输出时忘了运行

y_pred=(y_pred>0.5)

这一行，查看y_pred的数据类型是%f,而y_test的数据类型是%d

解决方法

运行

y_pred=(y_pred>0.5)

bug解决