Time2GraphRework

Improvement of shapelet extraction for Time2Graph

AQOURSNet: Time2Graph Rework

Ziyuan Chen, Zhirong Chen | July 2022
Summer Research @ Yang Yang Lab, Zhejiang University

This work is protected under the MIT License.
Copyright (c) 2022 Ziyuan Chen & Zhirong Chen unless otherwise noted.

Autoencoded
Quantification
Of
Unsupervised
Representative
Shapelets

Running the Program

$ python mainczr.py

• WARNING: the condensed ts2vec module has not been thoroughly tested. Clone yuezhihan/ts2vec under the same folder in case it fails.

Model Pipeline

0. Data preparation
1. Time series ---extract---> Shapelets
2. Time series & shapelets ---embed---> Series embedding
3. Series embedding ---construct---> Graph
4. Graph ---embed---> Graph embedding
5. Graph embedding ---predict---> Predicted classes

Folder Structure

• ref_papers - papers associated with shapelets providing essential background knowledge
  • traditional - evolution path since 2009: Logical, ST, Unsupervised, FS, LS, Random, Forest, Random forest, etc.
  • neural_network - DNNs enter the game! Dynamic, Adversarial, Adv. Dynamic, ShapeNet, BSPCover, etc. (Review)
• cached_programs - historical versions and experiments of KMeans, SVM, MLP, ResNet, Time2Vec, hierarchies, etc.
  • WARNING: Codes in the cache may not run properly since they are not optimized for environment compatibility.
• affiliated_licenses - LICENSEs for code segments from yuezhihan, subhadarship, DynamicTimeWarping, and pyg-team@GitHub.
• ucr_dataset - a neatly formatted version of the UCR Dataset in .npz format
  • The numpy arrays contained in each file have keys train_data, train_label, test_data, test_label
  • *_data has shape (num_samples, num_features), *_label has shape (num_samples,)

Call Hierarchy

• mainczr.py
  • utils.py (#0: ucr_dataset, #3 wrapper: GraphDataset)
  • construct_graph.py (#1 wrapper: extract_shapelets, #2: embed_series, #3: adjacency_matrix)
    • ts2vec.py (#1 optional enhancement: TS2Vec)
    • kmeans.py (#1: kmeans)
    • dtw
  • network.py (#4: GAT, #5: MultilayerPerceptron & FCResidualNetwork)
    • xgboost

Credits

• Original Time2Graph model by Cheng et al., 2020 & Cheng et al., 2021, code inspired by petecheng@GitHub
• Time2Vec algorithm (essence of AQOURS) by Kazemi et al., 2019, code adapted from yuezhihan@GitHub
  • Copyright (c) 2022 Zhihan Yue
• KMeans acceleration on PyTorch by subhadarship@GitHub, with adaptations
  • Copyright (c) 2020 subhadarshi
• DTW algorithm implemented by DynamicTimeWarping@GitHub
  • Copyright (c) 2019 Toni Giorgino
• GAT structure by Veličković et al., 2017, code inspired by Stanford CS224w, pyg-team@GitHub, DGraphXinye@GitHub
  • Copyright (c) 2021 Matthias Fey, Jiaxuan You (pyg-team)
• MLP & ResNet (prediction tail) structures inspired by Wang et al., 2017
• Data from the UCR Time Series Classification Archive, as documented in Dau et al., 2018
• Python modules: torch, torch_geometric, kmeans, dtw, xgboost, sklearn