多个模型分别预测

# 导包
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC, LinearSVC
from sklearn.ensemble import RandomForestClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.linear_model import Perceptron
from sklearn.linear_model import SGDClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.metrics import confusion_matrix# 生成用来记录评价指标的dataframe kpi
cols_list&＃61;[&＃39;ACC&＃39;,&＃39;TPR&＃39;,&＃39;TNR&＃39;]
kpi&＃61;pd.DataFrame(columns&＃61;cols_list)# 多个机器学习模型
logreg &＃61; LogisticRegression()
svc &＃61; SVC()
linear_svc &＃61; LinearSVC()
knn &＃61; KNeighborsClassifier(n_neighbors &＃61; 3)
gaussian &＃61; GaussianNB()
decision_tree &＃61; DecisionTreeClassifier()
random_forest &＃61; RandomForestClassifier(n_estimators&＃61;100) # 构造树的个数ways&＃61;[logreg,svc,linear_svc,knn,gaussian,decision_tree,random_forest]# 预测
for way in ways:way.fit(X_train, Y_train)Y_pred &＃61; way.predict(X_test)# 通过混淆矩阵计算评价指标TP&＃61;confusion_matrix(Y_test,Y_pred)[0][0]FN&＃61;confusion_matrix(Y_test,Y_pred)[0][1]FP&＃61;confusion_matrix(Y_test,Y_pred)[1][0]TN&＃61;confusion_matrix(Y_test,Y_pred)[1][1]new&＃61;pd.DataFrame(columns&＃61;cols_list,index&＃61;[str(way)])new[&＃39;ACC&＃39;]&＃61;round(way.score(X_test, Y_test) * 100, 2)new[&＃39;TPR&＃39;]&＃61;TP/(TP&＃43;FN) # 灵敏度new[&＃39;TNR&＃39;]&＃61;TN/(FP&＃43;TN) # 特异度kpi&＃61;kpi.append(new,ignore_index&＃61;False)