机器学习不同分类模型精度对比实现报错求助

项目说明 #

该项目用于练习分类模型技能，实现决策树、Logistic Regression、SVM等不同算法的精度对比，分析各模型的表现特性。

原问题代码 #

import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn import svm
from sklearn import metrics

def main():
    x, y = loadData()
    x_train, x_test, y_train, y_test = trainTest(x,y)
    logisticAccuracy = logisticRegressionModel(x_train, x_test, y_train, y_test)
    dectreeAccuracy = decisionTreeModel(x_train, x_test, y_train, y_test)
    svmAccuracy = svmModel(x_train, x_test, y_train, y_test)
    printComparisonResults(logisticAccuracy,dectreeAccuracy,svmAccuracy)

def loadData():
    df = pd.read_csv('D:\Tutorials\Resources\creditcard.csv')
    x = df[['Time','V1','V2','V3','V5','V6','V7','V8','V9','V10','V11','V12',\
    'V13','V14','V5','V16','V17','V18','V19','V20','V21','V22','V23',\
    'V24','V25','V26','V27', 'V28','Amount']]
    y = df['Class']

    return x, y

def trainTest(x,y):
    x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.25, random_state=0)
    return x_train, x_test, y_train, y_test

# LogisticRegression model
def logisticRegressionModel(x_train, x_test, y_train, y_test):
    logistic_reg = LogisticRegression().fit(x_train, y_train)
    y_pred = logistic_reg.predict(x_test)

    #evaluate accuracy of our model
    acc = metrics.accuracy_score(y_test, y_pred)

    return acc

def decisionTreeModel(x_train, x_test, y_train, y_test):
    dec_tree = DecisionTreeClassifier()
    dec_tree = dec_tree.fit(x_train, y_train)
    y_pred = dec_tree.predict(x_test)

    #evaluate model
    acc = metrics.accuracy_score(y_test, y_pred)

    return acc

def svmModel(x_train, x_test, y_train, y_test):
    svm_model = svm.SVC(kernel='linear')
    svm_model = svm_model.fit(x_train, y_train)
    y_pred = svm_model.predict(y_test)

    #evaluate model
    acc = metrics.accuracy_score(y_test, y_pred)

    return acc

def printComparisonResults(logisticAccuracy,dectreeAccuracy,svmAccuracy):
    print("Printing Results")

    lr = "Logistic Regression"
    dt = "Decision Tree"
    sv = "SVM is Accurate"

    if logisticAccuracy > (dectreeAccuracy and svmAccuracy):
        print(lr, "is accurate than", dt, "and", sv)
    elif dectreeAccuracy > (logisticAccuracy and svmAccuracy):
        print(dt, "is accurate than", lr, "and", sv)
    else:
        print(sv, "is accurate than", lr, "and", dt)

    print("Done")

if name == 'main': main()

首次运行报错 #

ValueError: Expected 2D array, got 1D array instead:
array=[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.].
Reshape your data either using array.reshape(-1, 1) if your data has a single feature or array.reshape(1, -1) if it contains a single sample.

错误修改尝试与二次报错 #

尝试修改特征数组为x = np.array(df[[...]].reshape((1,-1))后，出现新报错：

ValueError: Found input variables with inconsistent numbers of samples: [1, 2470]

问题原因与修复方案 #

核心错误1：SVM预测参数传错 #

svmModel函数中调用predict方法时，错误传入了1维的标签数组y_test，应该传入测试集特征x_test，直接导致模型收到1维数组报错。

错误2：特征reshape逻辑错误 #

原特征数组本身是符合模型要求的2维结构（样本数*特征数），强行reshape为(1,-1)会把所有样本压缩为1个样本，导致特征样本数和标签样本数不匹配，触发第二个报错，直接删掉这行修改代码即可。

其他需要修复的问题： #

• 特征选择错误：loadData函数中特征列重复选择了V5，删除重复的V5即可
• 逻辑回归收敛问题：LogisticRegression初始化时添加max_iter=1000参数，避免训练不收敛报警告
• 精度比较逻辑错误：原判断logisticAccuracy > (dectreeAccuracy and svmAccuracy)逻辑错误，正确写法为logisticAccuracy > dectreeAccuracy and logisticAccuracy > svmAccuracy
• 程序入口错误：原入口判断应为if __name__ == '__main__':，缺少双下划线会导致程序无法正常启动

修复后的关键代码片段 #

# 修正SVM预测参数
def svmModel(x_train, x_test, y_train, y_test):
    svm_model = svm.SVC(kernel='linear')
    svm_model = svm_model.fit(x_train, y_train)
    y_pred = svm_model.predict(x_test) # 这里修改为传入x_test
    acc = metrics.accuracy_score(y_test, y_pred)
    return acc

# 修正逻辑回归初始化
def logisticRegressionModel(x_train, x_test, y_train, y_test):
    logistic_reg = LogisticRegression(max_iter=1000).fit(x_train, y_train)
    y_pred = logistic_reg.predict(x_test)
    acc = metrics.accuracy_score(y_test, y_pred)
    return acc

# 修正精度比较逻辑
def printComparisonResults(logisticAccuracy,dectreeAccuracy,svmAccuracy):
    print("Printing Results")
    lr = "Logistic Regression"
    dt = "Decision Tree"
    sv = "SVM"
    if logisticAccuracy > dectreeAccuracy and logisticAccuracy > svmAccuracy:
        print(lr, "is more accurate than", dt, "and", sv)
    elif dectreeAccuracy > logisticAccuracy and dectreeAccuracy > svmAccuracy:
        print(dt, "is more accurate than", lr, "and", sv)
    else:
        print(sv, "is more accurate than", lr, "and", dt)
    print("Done")

# 修正程序入口
if __name__ == '__main__': 
    main()

内容的提问来源于stack exchange，提问作者Charles