no message

shenweichen · Nov 3, 2022 · 3c64524 · 3c64524
1 parent 2405b38
commit 3c64524
Show file tree

Hide file tree

Showing 11 changed files with 56 additions and 44 deletions.
diff --git a/deepctr/layers/__init__.py b/deepctr/layers/__init__.py
@@ -1,11 +1,11 @@
 import tensorflow as tf
 
 from .activation import Dice
-from .core import DNN, LocalActivationUnit, PredictionLayer, RegulationLayer
+from .core import DNN, LocalActivationUnit, PredictionLayer, RegulationModule
 from .interaction import (CIN, FM, AFMLayer, BiInteractionPooling, CrossNet, CrossNetMix,
                           InnerProductLayer, InteractingLayer,
                           OutterProductLayer, FGCNNLayer, SENETLayer, BilinearInteraction,
-                          FieldWiseBiInteraction, FwFMLayer, FEFMLayer, BridgeLayer)
+                          FieldWiseBiInteraction, FwFMLayer, FEFMLayer, BridgeModule)
 from .normalization import LayerNormalization
 from .sequence import (AttentionSequencePoolingLayer, BiasEncoding, BiLSTM,
                        KMaxPooling, SequencePoolingLayer, WeightedSequenceLayer,
@@ -28,7 +28,7 @@
                   'SequencePoolingLayer': SequencePoolingLayer,
                   'AttentionSequencePoolingLayer': AttentionSequencePoolingLayer,
                   'CIN': CIN,
-                  'RegulationLayer': RegulationLayer,
+                  'RegulationLayer': RegulationModule,
                   'InteractingLayer': InteractingLayer,
                   'LayerNormalization': LayerNormalization,
                   'BiLSTM': BiLSTM,
@@ -50,5 +50,5 @@
                   'FEFMLayer': FEFMLayer,
                   'reduce_sum': reduce_sum,
                   'PositionEncoding': PositionEncoding,
-                  'BridgeLayer': BridgeLayer
+                  'BridgeLayer': BridgeModule
                   }
diff --git a/deepctr/layers/core.py b/deepctr/layers/core.py
@@ -267,7 +267,7 @@ def get_config(self, ):
         return dict(list(base_config.items()) + list(config.items()))
 
 
-class RegulationLayer(Layer):
+class RegulationModule(Layer):
     """Regulation module used in EDCN.
 
       Input shape
@@ -280,17 +280,15 @@ class RegulationLayer(Layer):
         - **tau** : Positive float, the temperature coefficient to control
         distribution of field-wise gating unit.
 
-        - **seed** : A Python integer to use as random seed.
-
       References
         - [Enhancing Explicit and Implicit Feature Interactions via Information Sharing for Parallel Deep CTR Models.](https://dlp-kdd.github.io/assets/pdf/DLP-KDD_2021_paper_12.pdf)
     """
 
     def __init__(self, tau=1.0, **kwargs):
         if tau == 0:
-            raise ValueError("RegulationLayer tau can not be zero.")
+            raise ValueError("RegulationModule tau can not be zero.")
         self.tau = 1.0 / tau
-        super(RegulationLayer, self).__init__(**kwargs)
+        super(RegulationModule, self).__init__(**kwargs)
 
     def build(self, input_shape):
         self.field_size = int(input_shape[1])
@@ -301,7 +299,7 @@ def build(self, input_shape):
             name=self.name + '_field_weight')
 
         # Be sure to call this somewhere!
-        super(RegulationLayer, self).build(input_shape)
+        super(RegulationModule, self).build(input_shape)
 
     def call(self, inputs, **kwargs):
 
@@ -318,6 +316,6 @@ def compute_output_shape(self, input_shape):
 
     def get_config(self):
         config = {'tau': self.tau}
-        base_config = super(RegulationLayer, self).get_config()
+        base_config = super(RegulationModule, self).get_config()
         base_config.update(config)
         return base_config
diff --git a/deepctr/layers/interaction.py b/deepctr/layers/interaction.py
@@ -1493,8 +1493,8 @@ def get_config(self):
         return config
 
 
-class BridgeLayer(Layer):
-    """BridgeLayer layer used in EDCN
+class BridgeModule(Layer):
+    """Bridge Module used in EDCN
 
       Input shape
         - A list of two 2D tensor with shape: ``(batch_size, units)``.
@@ -1506,37 +1506,32 @@ class BridgeLayer(Layer):
         - **bridge_type**: The type of bridge interaction, one of 'pointwise_addition', 'hadamard_product', 'concatenation', 'attention_pooling'
 
         - **activation**: Activation function to use.
-       
-        - **l2_reg**: float between 0 and 1. L2 regularizer strength applied to the kernel weights matrix.
-
-        - **seed**: A Python integer to use as random seed.
 
       References
         - [Enhancing Explicit and Implicit Feature Interactions via Information Sharing for Parallel Deep CTR Models.](https://dlp-kdd.github.io/assets/pdf/DLP-KDD_2021_paper_12.pdf)
 
     """
 
-    def __init__(self, bridge_type='hadamard_product', activation='relu', l2_reg=0, seed=1024, **kwargs):
+    def __init__(self, bridge_type='hadamard_product', activation='relu', **kwargs):
         self.bridge_type = bridge_type
         self.activation = activation
-        self.l2_reg = l2_reg
-        self.seed = seed
 
-        super(BridgeLayer, self).__init__(**kwargs)
+        super(BridgeModule, self).__init__(**kwargs)
 
     def build(self, input_shape):
         if not isinstance(input_shape, list) or len(input_shape) < 2:
             raise ValueError(
-                'A `BridgeLayer` layer should be called '
-                'on a list of at least 2 inputs')
+                'A `BridgeModule` layer should be called '
+                'on a list of 2 inputs')
 
         self.dnn_dim = int(input_shape[0][-1])
+        if self.bridge_type == "concatenation":
+            self.dense = Dense(self.dnn_dim, self.activation)
+        elif self.bridge_type == "attention_pooling":
+            self.dense_x = DNN([self.dnn_dim, self.dnn_dim], self.activation, output_activation='softmax')
+            self.dense_h = DNN([self.dnn_dim, self.dnn_dim], self.activation, output_activation='softmax')
 
-        self.dense = Dense(self.dnn_dim, self.activation)
-        self.dense_x = DNN([self.dnn_dim, self.dnn_dim], self.activation, output_activation='softmax')
-        self.dense_h = DNN([self.dnn_dim, self.dnn_dim], self.activation, output_activation='softmax')
-
-        super(BridgeLayer, self).build(input_shape)  # Be sure to call this somewhere!
+        super(BridgeModule, self).build(input_shape)  # Be sure to call this somewhere!
 
     def call(self, inputs, **kwargs):
         x, h = inputs
@@ -1545,7 +1540,7 @@ def call(self, inputs, **kwargs):
         elif self.bridge_type == "hadamard_product":
             return x * h
         elif self.bridge_type == "concatenation":
-            return self.dense(tf.concat(inputs, axis=-1))
+            return self.dense(tf.concat([x, h], axis=-1))
         elif self.bridge_type == "attention_pooling":
             a_x = self.dense_x(x)
             a_h = self.dense_h(h)
@@ -1555,12 +1550,10 @@ def compute_output_shape(self, input_shape):
         return (None, self.dnn_dim)
 
     def get_config(self):
-        base_config = super(BridgeLayer, self).get_config().copy()
+        base_config = super(BridgeModule, self).get_config().copy()
         config = {
             'bridge_type': self.bridge_type,
-            'l2_reg': self.l2_reg,
-            'activation': self.activation,
-            'seed': self.seed
+            'activation': self.activation
         }
         config.update(base_config)
         return config
diff --git a/deepctr/models/edcn.py b/deepctr/models/edcn.py
@@ -10,16 +10,16 @@
 from tensorflow.python.keras.models import Model
 
 from ..feature_column import build_input_features, get_linear_logit, input_from_feature_columns
-from ..layers.core import PredictionLayer, DNN, RegulationLayer
-from ..layers.interaction import CrossNet, BridgeLayer
+from ..layers.core import PredictionLayer, DNN, RegulationModule
+from ..layers.interaction import CrossNet, BridgeModule
 from ..layers.utils import add_func, concat_func
 
 
 def EDCN(linear_feature_columns,
          dnn_feature_columns,
          cross_num=2,
          cross_parameterization='vector',
-         bridge_type='hadamard_product',
+         bridge_type='concatenation',
          tau=1.0,
          l2_reg_linear=1e-5,
          l2_reg_embedding=1e-5,
@@ -31,6 +31,7 @@ def EDCN(linear_feature_columns,
          dnn_activation='relu',
          task='binary'):
     """Instantiates the Enhanced Deep&Cross Network architecture.
+
     :param linear_feature_columns: An iterable containing all the features used by linear part of the model.
     :param dnn_feature_columns: An iterable containing all the features used by deep part of the model.
     :param cross_num: positive integet,cross layer number
@@ -63,25 +64,24 @@ def EDCN(linear_feature_columns,
         features, dnn_feature_columns, l2_reg_embedding, seed, support_dense=False)
 
     emb_input = concat_func(sparse_embedding_list, axis=1)
-    deep_in = RegulationLayer(tau)(emb_input)
-    cross_in = RegulationLayer(tau)(emb_input)
+    deep_in = RegulationModule(tau)(emb_input)
+    cross_in = RegulationModule(tau)(emb_input)
 
     field_size = len(sparse_embedding_list)
     embedding_size = int(sparse_embedding_list[0].shape[-1])
     cross_dim = field_size * embedding_size
 
     for i in range(cross_num):
-
         cross_out = CrossNet(1, parameterization=cross_parameterization,
                              l2_reg=l2_reg_cross)(cross_in)
         deep_out = DNN([cross_dim], dnn_activation, l2_reg_dnn,
                        dnn_dropout, dnn_use_bn, seed=seed)(deep_in)
         print(cross_out, deep_out)
-        bridge_out = BridgeLayer(bridge_type)([cross_out, deep_out])
+        bridge_out = BridgeModule(bridge_type)([cross_out, deep_out])
         if i + 1 < cross_num:
             bridge_out_list = Reshape([field_size, embedding_size])(bridge_out)
-            deep_in = RegulationLayer(tau)(bridge_out_list)
-            cross_in = RegulationLayer(tau)(bridge_out_list)
+            deep_in = RegulationModule(tau)(bridge_out_list)
+            cross_in = RegulationModule(tau)(bridge_out_list)
 
     stack_out = Concatenate()([cross_out, deep_out, bridge_out])
     final_logit = Dense(1, use_bias=False)(stack_out)

diff --git a/docs/pics/EDCN.jpg b/docs/pics/EDCN.jpg
diff --git a/docs/source/Features.md b/docs/source/Features.md
@@ -413,6 +413,17 @@ information routing across tasks in a general setup.
 
 [Tang H, Liu J, Zhao M, et al. Progressive layered extraction (ple): A novel multi-task learning (mtl) model for personalized recommendations[C]//Fourteenth ACM Conference on Recommender Systems. 2020.](https://dl.acm.org/doi/10.1145/3383313.3412236)
 
+### EDCN(Enhancing Explicit and Implicit Feature Interactions DCN)
+
+EDCN introduces two advanced modules, namelybridge moduleandregulation module, which work collaboratively tocapture the layer-wise interactive signals and learn discriminativefeature distributions for each hidden layer of the parallel networks.
+
+[**EDCN Model API**](./deepctr.models.edcn.html)
+
+![EDCN](../pics/EDCN.png)
+
+[Chen B, Wang Y, Liu Z, et al. Enhancing explicit and implicit feature interactions via information sharing for parallel deep ctr models[C]//Proceedings of the 30th ACM International Conference on Information & Knowledge Management. 2021: 3757-3766.](https://dlp-kdd.github.io/assets/pdf/DLP-KDD_2021_paper_12.pdf)
+
+
 ## Layers
 
 The models of deepctr are modular, so you can use different modules to build your own models.

diff --git a/docs/source/History.md b/docs/source/History.md
@@ -1,4 +1,5 @@
 # History
+- 11/05/2022 : [v0.9.3](https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.3) released.Add [EDCN]().
 - 10/15/2022 : [v0.9.2](https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.2) released.Support python `3.9`,`3.10`.
 - 06/11/2022 : [v0.9.1](https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.1) released.Improve compatibility with tensorflow `2.x`.
 - 09/03/2021 : [v0.9.0](https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.0) released.Add multitask learning models:[SharedBottom](./Features.html#sharedbottom),[ESMM](./Features.html#esmm-entire-space-multi-task-model),[MMOE](./Features.html#mmoe-multi-gate-mixture-of-experts) and [PLE](./Features.html#ple-progressive-layered-extraction). [running example](./Examples.html#multitask-learning-mmoe)

diff --git a/docs/source/Models.rst b/docs/source/Models.rst
@@ -30,5 +30,6 @@ DeepCTR Models API
    ESMM<deepctr.models.multitask.esmm>
    MMOE<deepctr.models.multitask.mmoe>
    PLE<deepctr.models.multitask.ple>
+   EDCN<deepctr.models.edcn>
 
 
diff --git a/docs/source/deepctr.models.edcn.rst b/docs/source/deepctr.models.edcn.rst
@@ -0,0 +1,7 @@
+deepctr.models.edcn module
+=========================
+
+.. automodule:: deepctr.models.edcn
+    :members:
+    :no-undoc-members:
+    :no-show-inheritance:
diff --git a/docs/source/deepctr.models.rst b/docs/source/deepctr.models.rst
@@ -11,6 +11,7 @@ Submodules
    deepctr.models.ccpm
    deepctr.models.dcn
    deepctr.models.dcnmix
+   deepctr.models.edcn
    deepctr.models.deepfm
    deepctr.models.dien
    deepctr.models.din

diff --git a/docs/source/index.rst b/docs/source/index.rst
@@ -42,12 +42,12 @@ You can read the latest code and related projects
 
 News
 -----
+11/05/2022 : Add `EDCN` . `Changelog <https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.3>`_
+
 10/15/2022 : Support python `3.9`,`3.10`. `Changelog <https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.2>`_
 
 06/11/2022 : Improve compatibility with tensorflow `2.x`. `Changelog <https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.1>`_
 
-09/03/2021 : Add multitask learning models: `SharedBottom <./Features.html#sharedbottom>`_ , `ESMM <./Features.html#esmm-entire-space-multi-task-model>`_ , `MMOE <./Features.html#mmoe-multi-gate-mixture-of-experts>`_ , `PLE <./Features.html#ple-progressive-layered-extraction>`_ .  `running example <./Examples.html#multitask-learning-mmoe>`_ `Changelog <https://github.com/shenweichen/DeepCTR/releases/tag/v0.9.0>`_
-
 DisscussionGroup
 -----------------------