Question

导入库

Positioned.fill(
    child: Image.asset('images/back.png', fit: BoxFit.cover),
    right: 0,
    left: 0
);

读取数据

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import sklearn 
from sklearn import preprocessing
import seaborn as sns
%matplotlib inline

将数据分为x和y：

 df =pd.read_csv('./EngineeredData_2.csv')
    df =df.dropna()

将y整形为整数：

X= df.drop (['Week','Div', 'Date', 'HomeTeam', 'AwayTeam','HTHG', 'HTAG','HTR', 
            'FTAG', 'FTHG','HGKPP', 'AGKPP', 'FTR'], axis =1)

将数据拆分为训练并进行测试：

 L = preprocessing.LabelEncoder ()
    matchresults = L.fit_transform (list (df['FTR']))
    y =list(matchresults)

导入类

from sklearn.model_selection import train_test_split
X_tng,X_tst, y_tng, y_tst =train_test_split (X, y, test_size = 50, shuffle=False)
X_tng.head()

实例化模型

from sklearn.linear_model import LogisticRegression

使模型适合数据

logreg = LogisticRegression ()

预测测试数据 y_pred = logreg.predict（X_tst）

 logreg.fit (X_tng, y_tng)

准确度达到100％有意义吗？

Answer 1

问题是您无意中删除了所有功能，而仅将目标值保留在x中。因此，您试图用目标值本身来解释目标值，这当然会给您100％的准确性。您将功能列定义为：

X= df.drop (['Week','Div', 'Date', 'HomeTeam', 'AwayTeam','HTHG', 'HTAG','HTR', 
            'FTAG', 'FTHG','HGKPP', 'AGKPP', 'FTR'], axis =1)

但是您应该将它们定义为：

X= df.drop('FTR', axis =1)

为什么我的逻辑回归模型获得100％的准确性？

1 个答案: