使用sklearn 神经网络训练模型

最新推荐文章于 2024-07-02 11:31:07 发布

XH生信ML笔记

最新推荐文章于 2024-07-02 11:31:07 发布

阅读量3.5k

点赞数 2

分类专栏：神经网络机器学习

本文链接：https://blog.csdn.net/weixin_44022515/article/details/103286222

版权

使用sklearn中神经网络模型对数据集进行训练和验证。
数据见本人博客“随机森林模型预测和交叉验证“

Neural Network trainning

>>> from sklearn.model_selection import train_test_split
>>> from sklearn.model_selection import cross_val_score
>>> Xtrain,Xtest,Ytrain,Ytest = train_test_split(allmatrix,target,test_size=0.3,random_state=420)
>>> from sklearn.neural_network import MLPClassifier
>>> dnn = MLPClassifier(hidden_layer_sizes=(50,50),random_state = 420) # 使用两层，每层50个节点
>>> dnn = dnn.fit(Xtrain,Ytrain)
>>> pred_dnn = dnn.predict(Xtest)
>>> score_dnn = ACCS(Ytest, pred_dnn)
>>> print("dnn training with laryer (50, 50) test score: {}".format(score_dnn))
dnn training with laryer (50, 50) test score: 0.8470485709712415

准确率同随机森林类似

dnn cv=5 cross_val交叉验证

>>> dnn_val=cross_val_score(dnn,allmatrix,target,cv=5)
>>> dnn_val
array([0.83468495, 0.83486287, 0.87552022, 0.81869598, 0.83283534])
>>> dnn_val.mean()
0.8393198731730974

归一化后的数据训练

>>> from sklearn.preprocessing import MinMaxScaler
>>> scaler = MinMaxScaler()
>>> X_mm=scaler.fit_transform(allmatrix)
>>> dnn_val_mm=cross_val_score(dnn,X_mm, target, cv