Question

我试图根据其他参数预测，数据为24个输入和1个输出（持续595天）。

我已经尝试创建具有 10倍交叉验证的神经网络，但是它给我带来了30％至15％的训练错误和40％的测试错误。

def create_model():
    model = Sequential()
    # Adding the input layer
    model.add(Dense(24, kernel_initializer='normal', activation='relu', input_shape=(24,)))
    # Adding the hidden layer
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(50, kernel_initializer='normal', activation='relu'))
    model.add(Dense(1))
    # Compiling the RNN
    model.compile(optimizer='adam', loss='mean_absolute_percentage_error')

    return model

kf = KFold(n_splits = 10, shuffle = True)
Density = create_model()

不知道如何最小化错误？还是有回归函数？

有用于非线性回归的scikit-learn / keras函数吗？

0 个答案: