向 skflow 添加正则化器

Question

我最近从 tensorflow 切换到 skflow。在 tensorflow 中，我们会将 lambda*tf.nn.l2_loss(weights) 添加到我们的损失中。现在我在 skflow 中有以下代码：

def deep_psi(X, y):
    layers = skflow.ops.dnn(X, [5, 10, 20, 10, 5], keep_prob=0.5)
    preds, loss = skflow.models.logistic_regression(layers, y)
    return preds, loss

def exp_decay(global_step):
    return tf.train.exponential_decay(learning_rate=0.01,
                                      global_step=global_step,
                                      decay_steps=1000,
                                      decay_rate=0.005)

deep_cd = skflow.TensorFlowEstimator(model_fn=deep_psi,
                                    n_classes=2,
                                    steps=10000,
                                    batch_size=10,
                                    learning_rate=exp_decay,
                                    verbose=True,)

如何以及在何处添加正则化项？ Illia 暗示了一些事情 here 但我想不通。

Answer 1

你仍然可以添加额外的组件到损失中，你只需要从 dnn / logistic_regression 中检索权重并将它们添加到损失中：

def regularize_loss(loss, weights, lambda):
    for weight in weights:
        loss = loss + lambda * tf.nn.l2_loss(weight)
    return loss    


def deep_psi(X, y):
    layers = skflow.ops.dnn(X, [5, 10, 20, 10, 5], keep_prob=0.5)
    preds, loss = skflow.models.logistic_regression(layers, y)

    weights = []
    for layer in range(5): # n layers you passed to dnn
        weights.append(tf.get_variable("dnn/layer%d/linear/Matrix" % layer))
        # biases are also available at dnn/layer%d/linear/Bias
    weights.append(tf.get_variable('logistic_regression/weights'))

    return preds, regularize_loss(loss, weights, lambda)

```

注意，变量的路径可以是found here.

此外，我们想为所有具有变量的层（如 dnn、conv2d 或 fully_connected）添加正则化器支持，因此下周的 Tensorflow 构建应该会有一些东西像这样 dnn(.., regularize=tf.contrib.layers.l2_regularizer(lambda))。发生这种情况时，我会更新此答案。

向 skflow 添加正则化器

Adding regularizer to skflow

python

neural-network

deep-learning

tensorflow

skflow