Sklearn神经网络算法MLP

MLP，神经网络算法，又称为多层感知机神经网络算法。

截屏2020-05-27上午11.24.03.png

代码：

import pandas as pd
import numpy as np
from sklearn import metrics
from sklearn.svm import SVC
from sklearn.neural_network import MLPClassifier
from sklearn.neural_network import MLPRegressor

# MLP神经网络算法    
def mx_MLP(train_x, train_y):    
    #mx = MLPClassifier(solver='lbfgs', alpha=1e-5,hidden_layer_sizes=(5, 2), random_state=1) 
    mx = MLPClassifier() 
    mx.fit(train_x, train_y)
    return mx

#结果验证函数
def ai_acc_xed(df9,ky0=5,fgDebug=True):
    #1
    #ny_test,ny_pred=len(df9['y_test']),len(df9['y_pred'])
    ny_test=len(df9['y_test'])
    df9['ysub']=df9['y_test']-df9['y_pred']
    df9['ysub2']=np.abs(df9['ysub'])
    #2
    df9['y_test_div']=df9['y_test']
    df9.loc[df9['y_test'] == 0, 'y_test_div'] =0.00001
    df9['ysubk']=(df9['ysub2']/df9['y_test_div'])*100
    dfk=df9[df9['ysubk']<ky0]
    dsum=len(dfk['y_pred'])
    dacc=dsum/ny_test*100
    #
    #3
    if fgDebug:
        print('\nai_acc_xed')
        print(df9.head())
        y_test,y_pred=df9['y_test'],df9['y_pred']
        print('\nn_df9,{0},n_dfk,{1}'.format(ny_test,dsum))
        dmae=metrics.mean_absolute_error(y_test, y_pred)
        dmse=metrics.mean_squared_error(y_test, y_pred)
        drmse=np.sqrt(metrics.mean_squared_error(y_test, y_pred))
        print('acc-kok: {0:.2f}%, MAE:{1:.2f}, MSE:{2:.2f}, RMSE:{3:.2f}'.format(dacc,dmae,dmse,drmse))

    #
    #4
    return dacc
# 1
fs0 = 'dat/iris_'
print('\n#1 init, fs0,', fs0)
x_train = pd.read_csv(fs0 + 'xtrain.csv', index_col = False)
y_train = pd.read_csv(fs0 + 'ytrain.csv', index_col = False)
x_test = pd.read_csv(fs0 + 'xtest.csv', index_col = False)
y_test = pd.read_csv(fs0 + 'ytest.csv', index_col = False)
df9 = x_test.copy()


#2
print('\n#2 model')
mx = mx_MLP(x_train.values, y_train.values)

# 3
print('\n#3 predict')
y_pred = mx.predict(x_test.values)
df9['y_predsr'] = y_pred
df9['y_test'], df9['y_pred'] = y_test, y_pred
df9['y_pred'] = round(df9['y_predsr']).astype(int)

# 4
df9.to_csv('dat/iris_9.csv', index = False)
print('\n#4 d9')
print(df9.tail())
 
# 5
dacc = ai_acc_xed(df9, 1, False)
print('\n#5 mx:mx_sum, kok: {0:.2f}%'.format(dacc))

输出：

#4 d9
     x1   x2   x3   x4  y_predsr  y_test  y_pred
33  6.4  2.8  5.6  2.1         1       1       1
34  5.8  2.8  5.1  2.4         1       1       1
35  5.3  3.7  1.5  0.2         2       2       2
36  5.5  2.3  4.0  1.3         3       3       3
37  5.2  3.4  1.4  0.2         2       2       2

#5 mx:mx_sum, kok: 97.37%