Home | History | Annotate | Download | only in models 
      1 # Copyright 2016 The TensorFlow Authors. All Rights Reserved.
      2 #
      3 # Licensed under the Apache License, Version 2.0 (the "License");
      4 # you may not use this file except in compliance with the License.
      5 # You may obtain a copy of the License at
      6 #
      7 #     http://www.apache.org/licenses/LICENSE-2.0
      8 #
      9 # Unless required by applicable law or agreed to in writing, software
     10 # distributed under the License is distributed on an "AS IS" BASIS,
     11 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
     12 # See the License for the specific language governing permissions and
     13 # limitations under the License.
     14 # ==============================================================================
     15 """A model that places a hard decision tree embedding before a neural net."""
     16 from __future__ import absolute_import
     17 from __future__ import division
     18 from __future__ import print_function
     19 
     20 from tensorflow.contrib import layers
     21 from tensorflow.contrib.tensor_forest.hybrid.python import hybrid_model
     22 from tensorflow.contrib.tensor_forest.hybrid.python.layers import decisions_to_data
     23 from tensorflow.contrib.tensor_forest.hybrid.python.layers import fully_connected
     24 from tensorflow.python.ops import nn_ops
     25 from tensorflow.python.training import adagrad
     26 
     27 
     28 class HardDecisionsToDataThenNN(hybrid_model.HybridModel):
     29   """A model that treats tree inference as hard at test."""
     30 
     31   def __init__(self,
     32                params,
     33                device_assigner=None,
     34                optimizer_class=adagrad.AdagradOptimizer,
     35                **kwargs):
     36 
     37     super(HardDecisionsToDataThenNN, self).__init__(
     38         params,
     39         device_assigner=device_assigner,
     40         optimizer_class=optimizer_class,
     41         **kwargs)
     42 
     43     self.layers = [decisions_to_data.HardDecisionsToDataLayer(
     44         params, 0, device_assigner),
     45                    fully_connected.FullyConnectedLayer(
     46                        params, 1, device_assigner=device_assigner)]
     47 
     48   def _base_inference(self, data, data_spec=None, soft=False):
     49     if soft:
     50       inference_result = self.layers[0].soft_inference_graph(data)
     51     else:
     52       inference_result = self._do_layer_inference(self.layers[0], data)
     53 
     54     for layer in self.layers[1:]:
     55       inference_result = self._do_layer_inference(layer, inference_result)
     56 
     57     output_size = 1 if self.is_regression else self.params.num_classes
     58     output = layers.fully_connected(
     59         inference_result, output_size, activation_fn=nn_ops.softmax)
     60     return output
     61 
     62   def inference_graph(self, data, data_spec=None):
     63     """Returns the op that performs inference on a batch of data."""
     64 
     65     return nn_ops.softmax(
     66         self._base_inference(
     67             data, data_spec=data_spec, soft=True))
     68 
     69   # pylint: disable=unused-argument
     70   def training_inference_graph(self, data, data_spec=None):
     71     return self._base_inference(data, data_spec=data_spec, soft=False)
     72