Skip to content
forked from betterzhou/DECL

Code for paper "Denoising-Aware Contrastive Learning for Noisy Time Series" (IJCAI 2024)

License

Notifications You must be signed in to change notification settings

denoiser-ml/DECL

This branch is up to date with betterzhou/DECL:main.

Folders and files

NameName
Last commit message
Last commit date

Latest commit

a992902 · Nov 21, 2024

History

17 Commits
Nov 20, 2024
Nov 20, 2024
Nov 20, 2024
Nov 20, 2024
Nov 20, 2024
Nov 20, 2024
Nov 20, 2024
Nov 20, 2024
Nov 21, 2024
Jun 7, 2024
Nov 21, 2024
Nov 21, 2024
Nov 20, 2024
Nov 21, 2024
Nov 20, 2024
Nov 20, 2024

Repository files navigation

Denoising-Aware Contrastive Learning for Noisy Time Series (DECL)

1. Introduction

This repository contains code for the paper "Denoising-Aware Contrastive Learning for Noisy Time Series" (IJCAI 2024).

2. Usage

Requirements:

  • torch==1.7.1
  • python==3.7.16

See requirements.txt for details.

Datasets:

The code assumes that the data has been augmented and saved into the data folder.

For example, the folder of ./PTB/ contains all the raw data, the folder of ./PTB_denoise1/ contains all the denoised data using the denoiser m1, the folder of ./PTB_noisy1/ contains all the noise-enhanced data corresponding to denoiser m1, and the folder of ./PTB_Gaussian/ contains all the data induced with Gaussian noise.

The meta information of each time series sample is in the label.csv file.

Example:

Please modify the 'data_path' in the code to adapt to the path of your data folder.

If using a few labels for training, please split data twice: one for pre-training and one for linear evaluation.

  • python run_preprocess.py --trn_ratio 0.4 --val_ratio 0.2 --tes_ratio 0.4 --dataset PTB --seed 1
  • python main.py --experiment_description exp1 --run_description run1 --data_path './data/' --selected_dataset PTB --seed 1 --training_mode self_supervised --lr 1e-6
  • python run_preprocess.py --trn_ratio 0.2 --val_ratio 0.1 --tes_ratio 0.4 --dataset PTB --seed 1
  • python main.py --experiment_description exp1 --run_description run1 --data_path './data/' --selected_dataset PTB --seed 1 --training_mode train_linear --lr 1e-6

We suggest setting learning_rate to small values.

For research cooperation, please contact shuang.zhou@connect.polyu.hk

3. Citation

Please kindly cite the paper if you use the code or any resources in this repo:

@inproceedings{ijcai2024p624,
  title     = {Denoising-Aware Contrastive Learning for Noisy Time Series},
  author    = {Zhou, Shuang and Zha, Daochen and Shen, Xiao and Huang, Xiao and Zhang, Rui and Chung, Korris},
  booktitle = {Proceedings of the Thirty-Third International Joint Conference on
               Artificial Intelligence, {IJCAI-24}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Kate Larson},
  pages     = {5644--5652},
  year      = {2024},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2024/624},
  url       = {https://doi.org/10.24963/ijcai.2024/624},
}

About

Code for paper "Denoising-Aware Contrastive Learning for Noisy Time Series" (IJCAI 2024)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%