classification_report精度低

Question

我的逻辑回归的 classification_report 精度很低，但我不知道为什么

from sklearn.model_selection import train_test_split

features = ['gender','age','hypertension','heart_disease','avg_glucose_level',
            'bmi','smoking_status']
data1 = ['stroke']

X = df[features]
y = df[data1]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=42)

from sklearn.linear_model import LogisticRegression

#setting class weight to 'balance' will help with the data imbalance giving higher weights to the data with less samples
reg = LogisticRegression(penalty='l2',solver='sag' ,class_weight='balanced').fit(X_train, y_train)

y_pred = reg.predict(X_test)

train_pred = reg.predict(X_train)
test_pred = reg.predict(X_test)

from sklearn.metrics import classification_report

#train report
print (classification_report(y_train ,train_pred))

#test report
print (classification_report(y_test ,test_pred))

我尝试添加求解器方法并使用 max_iter 并将其设置为 1000，但没有任何改变