Awesome

MambaTS

The repo is the official implementation for the paper: MambaTS: Improved Selective State Space Models for Long-term Time Series Forecasting.

Key codes:

For the architecture design of MambaTS, please refer primarily to models/MambaTS.py.
For variable permutation training (VPT), please focus on the random_shuffle and unshuffle function in utils/masking.py.
For variable-aware scan along time (VAST), please refer mainly to layers/mamba_ssm/mixer2_seq_simple.py.

Recently, we've also released a repo tracking the latest developments in Mamba. If you're interested, you can check it out at Awesome-Mamba-Collection and enjoy it.

Usage

Install Python 3.11. For convenience, execute the following command.
```
pip install -r requirements.txt
```
For setting up the Mamba environment, please refer to https://github.com/state-spaces/mamba. Here is a simple instruction on Linux system,
```
pip install causal-conv1d>=1.2.0
pip install mamba-ssm
```
Prepare Data. You can obtain the well pre-processed datasets from public channel like [Google Drive] or [Tsinghua Cloud], Then place the downloaded data in the folder./dataset.
Train and evaluate model. We provide the experiment scripts for MambaTS under the folder ./scripts/. You can reproduce the experiment results as the following examples:
```
# long-term forecast for ETTm2 dataset
bash ./scripts/long_term_forecast/MambaTS_ETTm2.sh
```

Acknowledgement

This library is constructed based on the following repos:

Time-Series-Library: https://github.com/thuml/Time-Series-Library
Mamba: https://github.com/state-spaces/mamba

All the experiment datasets are public, and we obtain them from the following links:

Long-term Forecasting: https://github.com/thuml/Autoformer and https://github.com/thuml/iTransformer.

We also greatly appreciate the python-tsp for providing efficient solvers for the Asymmetric Traveling Salesperson Problem (ATSP).

We extend our sincere thanks for their excellent work and repositories!

Citation

If you find this repo useful, please consider citing our paper.

@article{cai2024mambats,
  title={MambaTS: Improved Selective State Space Models for Long-term Time Series Forecasting},
  author={Cai, Xiuding and Zhu, Yaoyao and Wang, Xueyao and Yao, Yu},
  journal={arXiv preprint arXiv:2405.16440},
  year={2024}
}