python-自适应提升算法的多种对比方法及混淆矩阵的计算

1.KNN算法

#coding:utf-8
from sklearn.neighbors import KNeighborsClassifier
import numpy as np
from sklearn.metrics import confusion_matrix # 导入计算混淆矩阵的包
import pandas as pd

X_train=pd.read_excel('D:\\xtrain.xlsx', header=None)
Y_train=pd.read_excel('D:\\ytrain.xlsx',  header=None)
X_test=pd.read_excel('D:\\xtest.xlsx', header=None)
Y_test =pd.read_excel('D:\\ytest.xlsx',  header=None)

knn = KNeighborsClassifier(n_neighbors=5)  # 获取KNN对象
knn.fit(X_train, Y_train)  # 训练模型

# 评估模型
y_pre = knn.predict(X_test)
score = knn.score(X_test, Y_test)  # 调用打分函数
print("test set predictions:\n{}".format(Y_test))
print("test set score:{:.2f}".format(score))
if score > 0.9:
    x_new = np.array([[5, 2.9, 1, 0.3]])
    print("x_new.shape:{}".format(x_new.shape))
    prediction = knn.predict(x_new)  # 预测
    print("prediction:{}".format(prediction))

else:
    print("used train or test data is not available !")

X=Y_test
Y=y_pre

C1= confusion_matrix(X,Y) #True_label 真实标签 shape=(n,1);T_predict1 预测标签 shape=(n,1)

print(C1)

2、朴素贝叶斯

from sklearn import naive_bayes
import pandas as pd
from sklearn.metrics import confusion_matrix # 导入计算混淆矩阵的包

X_train=pd.read_excel('D:\\xtrain.xlsx', header=None)
Y_train=pd.read_excel('D:\\ytrain.xlsx',  header=None)
X_test=pd.read_excel('D:\\xtest.xlsx', header=None)
Y_test =pd.read_excel('D:\\ytest.xlsx',  header=None)

#定义模型
nb = naive_bayes.GaussianNB()
nb.fit(X_train,Y_train)
nb.get_params()

#预测模型得分
sorce = nb.score(X_test,Y_test)
print("此模型得分为%s"%sorce)

# 预测数据，预测特征值
print(nb.predict(X_test))
# 打印真实特征值
print(Y_test)


X=Y_test
Y=nb.predict(X_test)

C1= confusion_matrix(X,Y) #True_label 真实标签 shape=(n,1);T_predict1 预测标签 shape=(n,1)

print(C1)

3、bp神经网络

from sklearn.model_selection import train_test_split
from sklearn.neural_network import MLPClassifier
import pandas as pd
from sklearn.metrics import confusion_matrix # 导入计算混淆矩阵的包

X_train=pd.read_excel('D:\\xtrain.xlsx', header=None)
Y_train=pd.read_excel('D:\\ytrain.xlsx',  header=None)
X_test=pd.read_excel('D:\\a.xlsx', header=None)
Y_test =pd.read_excel('D:\\b.xlsx',  header=None)

#定义模型
bp = MLPClassifier(max_iter=10000)
bp.fit(X_train,Y_train)
bp.get_params()

#预测模型得分
sorce = bp.score(X_test,Y_test)
print("此模型得分为%s"%sorce)
# 预测数据，预测特征值
print(bp.predict(X_test))

# 打印真实特征值
#print(Y_test)

X=Y_test
Y=bp.predict(X_test)
C1= confusion_matrix(X,Y) #True_label 真实标签 shape=(n,1);T_predict1 预测标签 shape=(n,1)

print(C1)