python - keras/scikit-学习 : using fit_generator() with cross validation

Question

是否可以使用 Keras's scikit-learn API与 fit_generator() 方法一起使用？还是使用另一种方法来生成用于训练的批处理？我正在使用 SciPy 的稀疏矩阵，在输入到 Keras 之前必须将其转换为 NumPy 数组，但由于内存消耗高，我无法同时转换它们。这是我生成批处理的函数:

def batch_generator(X, y, batch_size):
    n_splits = len(X) // (batch_size - 1)
    X = np.array_split(X, n_splits)
    y = np.array_split(y, n_splits)

    while True:
        for i in range(len(X)):
            X_batch = []
            y_batch = []
            for ii in range(len(X[i])):
                X_batch.append(X[i][ii].toarray().astype(np.int8)) # conversion sparse matrix -> np.array
                y_batch.append(y[i][ii])
            yield (np.array(X_batch), np.array(y_batch))

和带有交叉验证的示例代码:

from sklearn.model_selection import StratifiedKFold, GridSearchCV
from sklearn import datasets

from keras.models import Sequential
from keras.layers import Activation, Dense
from keras.wrappers.scikit_learn import KerasClassifier

import numpy as np


def build_model(n_hidden=32):
    model = Sequential([
        Dense(n_hidden, input_dim=4),
        Activation("relu"),
        Dense(n_hidden),
        Activation("relu"),
        Dense(3),
        Activation("sigmoid")
    ])
    model.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"])
    return model


iris = datasets.load_iris()
X = iris["data"]
y = iris["target"].flatten()

param_grid = {
    "n_hidden": np.array([4, 8, 16]),
    "nb_epoch": np.array(range(50, 61, 5))
}

model = KerasClassifier(build_fn=build_model, verbose=0)
skf = StratifiedKFold(n_splits=5).split(X, y) # this yields (train_indices, test_indices)

grid = GridSearchCV(model, param_grid, cv=skf, verbose=2, n_jobs=4)
grid.fit(X, y)

print(grid.best_score_)
print(grid.cv_results_["params"][grid.best_index_])

更详细地解释一下，它使用 param_grid 中超参数的所有可能组合来构建模型。然后在 StratifiedKFold 提供的训练测试数据拆分 (folds) 上对每个模型进行训练和测试。然后给定模型的最终分数是所有折叠的平均分数。

那么是否可以在实际拟合之前在上面的代码中插入一些预处理子步骤以转换数据(稀疏矩阵)？

我知道我可以编写自己的交叉验证生成器，但它必须生成索引，而不是真实数据!

Answer 1

实际上，您可以使用生成器将稀疏矩阵作为 Keras 的输入。这是我在以前的项目中使用的版本:

> class KerasClassifier(KerasClassifier):
>     """ adds sparse matrix handling using batch generator
>     """
>     
>     def fit(self, x, y, **kwargs):
>         """ adds sparse matrix handling """
>         if not issparse(x):
>             return super().fit(x, y, **kwargs)
>         
>         ############ adapted from KerasClassifier.fit   ######################   
>         if self.build_fn is None:
>             self.model = self.__call__(**self.filter_sk_params(self.__call__))
>         elif not isinstance(self.build_fn, types.FunctionType):
>             self.model = self.build_fn(
>                 **self.filter_sk_params(self.build_fn.__call__))
>         else:
>             self.model = self.build_fn(**self.filter_sk_params(self.build_fn))
> 
>         loss_name = self.model.loss
>         if hasattr(loss_name, '__name__'):
>             loss_name = loss_name.__name__
>         if loss_name == 'categorical_crossentropy' and len(y.shape) != 2:
>             y = to_categorical(y)
>         ### fit => fit_generator
>         fit_args = copy.deepcopy(self.filter_sk_params(Sequential.fit_generator))
>         fit_args.update(kwargs)
>         ############################################################
>         self.model.fit_generator(
>                     self.get_batch(x, y, self.sk_params["batch_size"]),
>                                         samples_per_epoch=x.shape[0],
>                                         **fit_args)                      
>         return self                               
> 
>     def get_batch(self, x, y=None, batch_size=32):
>         """ batch generator to enable sparse input """
>         index = np.arange(x.shape[0])
>         start = 0
>         while True:
>             if start == 0 and y is not None:
>                 np.random.shuffle(index)
>             batch = index[start:start+batch_size]
>             if y is not None:
>                 yield x[batch].toarray(), y[batch]
>             else:
>                 yield x[batch].toarray()
>             start += batch_size
>             if start >= x.shape[0]:
>                 start = 0
>   
>     def predict_proba(self, x):
>         """ adds sparse matrix handling """
>         if not issparse(x):
>             return super().predict_proba(x)
>             
>         preds = self.model.predict_generator(
>                     self.get_batch(x, None, self.sk_params["batch_size"]), 
>                                                val_samples=x.shape[0])
>         return preds