Tree-based Convolutional Neural Networks

Preface......Page 3
Contents......Page 5
Acronyms......Page 8
Deep Learning Background......Page 9
Incorporating Structural Information into Neural Architectures......Page 11
Proposed Tree-based Convolutional NNs......Page 12
Structure of the Book......Page 13
Refs......Page 14
Neuron & Multilayer Network......Page 16
Training Objectives......Page 19
Learning Neural Parameters......Page 21
Pretraining Neural Networks......Page 25
Neural Networks for NLP......Page 26
Neural Language Models......Page 27
Word Embeddings......Page 29
Convolutional Neural Network......Page 32
Recurrent Neural Network......Page 35
Recursive Neural Network......Page 37
Refs......Page 39
General Idea & Formula of TBCNN......Page 43
Difficulties in Designing TBCNN......Page 45
Intro......Page 47
Overview......Page 49
Representation Learning for AST Nodes......Page 50
Tree-based Convolutional Layer......Page 51
Dynamic Pooling......Page 52
Continuous Binary Tree Model......Page 53
Unsupervised Program Vector Representations......Page 54
Classifying Programs by Functionalities......Page 56
Detecting Bubble Sort......Page 58
Model Analysis......Page 59
Summary & Discussion......Page 60
Refs......Page 61
Sentence Modeling & Constituency Trees......Page 64
Recursively Representing Intermediate Nodes......Page 66
Constituency Tree-based Convolutional Layer......Page 68
Experiments......Page 69
Sentiment Analysis......Page 70
Question Classification......Page 73
Summary & Discussion......Page 74
Refs......Page 75
Dependency Trees......Page 77
Dependency Trees as Input......Page 78
Convolutional Layer......Page 79
Dynamic Pooling Layer......Page 80
Applying d-TBCNN to Sentence Matching......Page 81
Experiments......Page 83
Discriminative Sentence Modeling......Page 84
Sentence Matching......Page 85
Model Analysis......Page 87
Visualization......Page 89
Conclusion & Discussion......Page 91
Refs......Page 92
7 Conclusion & future Work......Page 94
Index......Page 98