def LeaveUserFeaturesHour():
X_train, X_test, y_train, y_test = data.Train_test_split()
predict = MLP_model(X_train, X_test, y_train, y_test)
y_test = y_test.reshape(1, -1)[0]
Evaluation.evalution(y_test, predict)
Evaluation.AUC(y_test, predict)
def pca():
X_train, X_test, y_train, y_test = data.Train_test_split()
pca = PCA(n_components=100)
X_train = pca.fit(X_train).transform(X_train)
pca1 = PCA(n_components=100)
X_test = pca1.fit(X_test).transform(X_test)
return X_train, X_test, y_train, y_test
temp = np.ones(x.shape[0] + 1)
temp[0:-1] = x
a = temp
for l in range(0, len(self.weights)):
a = self.activation(np.dot(a, self.weights[l]))
return a
'''
[2,2,1]
第一个2:表示 数据的纬度，因为是二维的，表示两个神经元，所以是2
第二个2：隐藏层数据纬度也是2，表示两个神经元
1：表示输入为一个神经元
tanh:表示用双曲函数里的tanh函数
'''
print("--------DNN---------")
X_train, X_test, y_train, y_test = data.Train_test_split()
nn = NeuralNetwork([100,2,1], 'tanh')
#X = np.array([[0, 0,0], [0, 1,0], [1, 0,0], [1, 1,0]])
#y = np.array([0, 1, 1, 0])
nn.fit(X_train, y_train)
predict = []
for i in X_test:
if nn.predict(i)[0] < 0 :predict.append(0)
else:predict.append(1)
#print(i,nn.predict(i))
y_test = y_test.reshape(1,-1)[0]
Evaluation.evalution(y_test,predict)
Evaluation.AUC(y_test,predict)