超参数调整以决定最佳神经网络

Posted 2023-03-12

【中文标题】超参数调整以决定最佳神经网络

【英文标题】：Hyperparameter tuning to decide optimal neural network

【发布时间】：2020-06-29 22:34:24

【问题描述】：

我想根据某些标准找出最佳神经网络。标准如下：

用一、二、三、四隐藏层 + 输出层测试 4 种架构

要测试的学习率：0.1,0.01,0.001

要测试的时期：10,50,100

输入尺寸 = 20

输出应该是显示每个组合的表格（36 行）。例如，一个隐藏层，lr = 0.1，epochs = 10，准确率是X。

请看下面我的代码：

#Function to create the model
def create_model(layers,learn_rate):
    model = Sequential()
    for i, nodes in enumerate(layers):
        if i==0:
            model.add(Dense(nodes),input_dim = 20,activation = 'relu')
        else:
            model.add(Dense(nodes),activation = 'relu')   

    model.add(Dense(units = 4,activation = 'softmax')) 

    model.compile(optimizer=adam(lr=learn_rate), loss='categorical_crossentropy',metrics=['accuracy'])
    return model

#Initialization of variables
#Here there are the four possible types of layers with the neurons in each.
layers = [[20], [40, 20], [45, 30, 15],[32,16,8,4]]
learn_rate = [0.1,0.01,0.001]
epochs = [10,50,100]

#GridSearchCV for hyperparameter tuning
from keras.wrappers.scikit_learn import KerasClassifier
from sklearn.model_selection import GridSearchCV
model = KerasClassifier(build_fn = create_model, verbose = 0)
param_grid = dict(layers = layers,learn_rate = learn_rate,epochs = epochs)
grid = GridSearchCV(estimator = model, param_grid = param_grid,cv = 3)
grid_result = grid.fit(train_x,train_y)

但是当我运行代码时，我收到以下错误：

RuntimeError: Cannot clone object <keras.wrappers.scikit_learn.KerasClassifier object at    0x000001AA272C7748>, as the constructor either does not set or modifies parameter layers

【问题讨论】：

正如错误消息所暗示的，您不能将层数用作KerasClassifier 中的超参数。您必须构建单独的模型并分别为每个模型运行网格搜索。

【参考方案1】：

无法克隆对象不是主要问题。这是模型生成器函数中另一个错误的结果。create_model() 中有一些语法错误。请查看输出中“克隆问题”之前出现的错误。 这是固定功能：

from keras import optimizers

def create_model(layers, learn_rate):
    model = Sequential()
    for i, nodes in enumerate(layers):
        if i==0:
            model.add(Dense(nodes,input_dim = 20,activation = 'relu'))
        else:
            model.add(Dense(nodes,activation = 'relu'))
    model.add(Dense(units = 4,activation = 'softmax')) 

    model.compile(optimizer=optimizers.adam(lr=learn_rate), loss='categorical_crossentropy',metrics=['accuracy']) 
    return model