Reference Style Fix (#4972)

examples/mnist_hierarchical_rnn.py

@@ -8,7 +8,7 @@ document vector is considered to preserve both the word-level and
 sentence-level structure of the context.
 
 # References
-    - [A Hierarchical Neural Autoencoder for Paragraphs and Documents](https://web.stanford.edu/~jurafsky/pubs/P15-1107.pdf)
+    - [A Hierarchical Neural Autoencoder for Paragraphs and Documents](https://arxiv.org/abs/1506.01057)
         Encodes paragraphs and documents with HRNN.
         Results have shown that HRNN outperforms standard
         RNNs and may play some role in more sophisticated generation tasks like

keras/layers/advanced_activations.py

@@ -65,7 +65,7 @@ class PReLU(Layer):
             set `shared_axes=[1, 2]`.
 
     # References
-        - [Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification](http://arxiv.org/pdf/1502.01852v1.pdf)
+        - [Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification](https://arxiv.org/abs/1502.01852)
     '''
     def __init__(self, init='zero', weights=None, shared_axes=None, **kwargs):
         self.supports_masking = True
@@ -124,7 +124,7 @@ class ELU(Layer):
         alpha: scale for the negative factor.
 
     # References
-        - [Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)](http://arxiv.org/pdf/1511.07289v1.pdf)
+        - [Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)](https://arxiv.org/abs/1511.07289v1)
     '''
     def __init__(self, alpha=1.0, **kwargs):
         self.supports_masking = True
@@ -228,7 +228,7 @@ class ThresholdedReLU(Layer):
         theta: float >= 0. Threshold location of activation.
 
     # References
-        - [Zero-Bias Autoencoders and the Benefits of Co-Adapting Features](http://arxiv.org/pdf/1402.3337.pdf)
+        - [Zero-Bias Autoencoders and the Benefits of Co-Adapting Features](http://arxiv.org/abs/1402.3337)
     '''
     def __init__(self, theta=1.0, **kwargs):
         self.supports_masking = True

keras/layers/convolutional.py

@@ -583,7 +583,7 @@ class Deconvolution2D(Convolution2D):
         `rows` and `cols` values might have changed due to padding.
 
     # References
-        [1] [A guide to convolution arithmetic for deep learning](https://arxiv.org/abs/1603.07285 "arXiv:1603.07285v1 [stat.ML]")
+        [1] [A guide to convolution arithmetic for deep learning](https://arxiv.org/abs/1603.07285v1)
         [2] [Transposed convolution arithmetic](http://deeplearning.net/software/theano_versions/dev/tutorial/conv_arithmetic.html#transposed-convolution-arithmetic)
         [3] [Deconvolutional Networks](http://www.matthewzeiler.com/pubs/cvpr2010/cvpr2010.pdf)
     '''

keras/layers/convolutional_recurrent.py

@@ -243,7 +243,7 @@ class ConvLSTM2D(ConvRecurrent2D):
 
     # References
         - [Convolutional LSTM Network: A Machine Learning Approach for
-        Precipitation Nowcasting](http://arxiv.org/pdf/1506.04214v1.pdf)
+        Precipitation Nowcasting](http://arxiv.org/abs/1506.04214v1)
         The current implementation does not include the feedback loop on the
         cells output
     '''

keras/layers/core.py

@@ -116,7 +116,7 @@ class SpatialDropout1D(Dropout):
         Same as input
 
     # References
-        - [Efficient Object Localization Using Convolutional Networks](https://arxiv.org/pdf/1411.4280.pdf)
+        - [Efficient Object Localization Using Convolutional Networks](https://arxiv.org/abs/1411.4280)
     '''
     def __init__(self, p, **kwargs):
         super(SpatialDropout1D, self).__init__(p, **kwargs)
@@ -154,7 +154,7 @@ class SpatialDropout2D(Dropout):
         Same as input
 
     # References
-        - [Efficient Object Localization Using Convolutional Networks](https://arxiv.org/pdf/1411.4280.pdf)
+        - [Efficient Object Localization Using Convolutional Networks](https://arxiv.org/abs/1411.4280)
     '''
     def __init__(self, p, dim_ordering='default', **kwargs):
         if dim_ordering == 'default':
@@ -202,7 +202,7 @@ class SpatialDropout3D(Dropout):
         Same as input
 
     # References
-        - [Efficient Object Localization Using Convolutional Networks](https://arxiv.org/pdf/1411.4280.pdf)
+        - [Efficient Object Localization Using Convolutional Networks](https://arxiv.org/abs/1411.4280)
     '''
     def __init__(self, p, dim_ordering='default', **kwargs):
         if dim_ordering == 'default':
@@ -875,7 +875,7 @@ class MaxoutDense(Layer):
         2D tensor with shape: `(nb_samples, output_dim)`.
 
     # References
-        - [Maxout Networks](http://arxiv.org/pdf/1302.4389.pdf)
+        - [Maxout Networks](http://arxiv.org/abs/1302.4389)
     '''
     def __init__(self, output_dim,
                  nb_feature=4,
@@ -1001,7 +1001,7 @@ class Highway(Layer):
         2D tensor with shape: `(nb_samples, input_dim)`.
 
     # References
-        - [Highway Networks](http://arxiv.org/pdf/1505.00387v2.pdf)
+        - [Highway Networks](http://arxiv.org/abs/1505.00387v2)
     '''
     def __init__(self,
                  init='glorot_uniform',

keras/layers/normalization.py

@@ -59,7 +59,7 @@ class BatchNormalization(Layer):
         Same shape as input.
 
     # References
-        - [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](http://jmlr.org/proceedings/papers/v37/ioffe15.pdf)
+        - [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://arxiv.org/abs/1502.03167)
     '''
     def __init__(self, epsilon=1e-3, mode=0, axis=-1, momentum=0.99,
                  weights=None, beta_init='zero', gamma_init='one',

keras/layers/recurrent.py

@@ -424,8 +424,8 @@ class GRU(Recurrent):
         dropout_U: float between 0 and 1. Fraction of the input units to drop for recurrent connections.
 
     # References
-        - [On the Properties of Neural Machine Translation: Encoder-Decoder Approaches](http://www.aclweb.org/anthology/W14-4012)
-        - [Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling](http://arxiv.org/pdf/1412.3555v1.pdf)
+        - [On the Properties of Neural Machine Translation: Encoder-Decoder Approaches](https://arxiv.org/abs/1409.1259)
+        - [Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling](http://arxiv.org/abs/1412.3555v1)
         - [A Theoretically Grounded Application of Dropout in Recurrent Neural Networks](http://arxiv.org/abs/1512.05287)
     '''
     def __init__(self, output_dim,