Official implementation of our NeurIPS 2024 paper Boosting Transferability and Discriminability for Time Series Domain Adaptation.

• We uncover the characteristics wherein temporal features and frequency features cannot be equally treated in transfer learning. Specifically, we observe that frequency features are more discriminative within a specific domain, while temporal features show better transferability across domains through empirical findings.

• We design ACON, which enhances UDA in three key aspects: a multi-period feature learning module to enhance the discriminability of frequency features, a temporal-frequency domain mutual learning module to enhance the discriminability of temporal features in the source domain and improve the transferability of frequency features in the target domain, and a domain adversarial learning module in temporal-frequency correlation subspace to further enhance transferability of features.

• Experiments conducted on eights time series datasets and five common applications verify the effectiveness.

• UCIHAR
• HHAR-P
• WISDM
• FD
• HHAR-D
• EMG
• PCL
• CAP

The train/test set split of HHAR_D, EMG, PCL, CAP:

• HHAR_D
• EMG
• PCL
• CAP

Data directory structure

.
└── data
    └── CAP
        ├── test_0.pt
        ├── test_1.pt
        ├── test_2.pt
        ├── test_3.pt
        ├── test_4.pt
        ├── train_0.pt
        ├── train_1.pt
        ├── train_2.pt
        ├── train_3.pt
        └── train_4.pt
    
    └── UCIHAR
      ......
    └── WISDM
      ......

For each dataset, we select 10 source-target domain pairs.

Detailed domain pairs are provided in data_model_configs.

Each experiment is repeated 5 times with different random seeds.

All bash scripts are provided in scripts.

To train a model on UCIHAR dataset:

CUDA_VISIBLE_DEVICES=0 python main.py \
 --experiment_description ACON \
 --run_description UCIHAR \
 --da_method ACON \
 --dataset UCIHAR \
 --num_runs 5 \
 --lr 0.01 \
 --cls_trade_off 1 \
 --domain_trade_off 1 \
 --entropy_trade_off 0.01 \
 --align_t_trade_off 1 \
 --align_s_trade_off 1

This repo is built on the pioneer works. We appreciate the following GitHub repos a lot for their valuable code base or datasets:

• RAINCOAT

• AdaTime

If you find this work helpful for your research, please kindly cite the following paper:

@inproceedings{liuboosting,
  title={Boosting Transferability and Discriminability for Time Series Domain Adaptation},
  author={Liu, Mingyang and Chen, Xinyang and Shu, Yang and Li, Xiucheng and Guan, Weili and Nie, Liqiang},
  booktitle={Annual Conference on Neural Information Processing Systems (NeurIPS)},
  year={2024}
}