visionts 0.1.0

Using a vision MAE for time series forecasting.

VisionTS

Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasters

📃 Paper | 🔍 About | 🔨 Setup | 🚀 Quick Start | 📊 Evaluation | 🔗 Citation

🔍 About

• We propose VisionTS, a time series forecasting (TSF) foundation model building from rich, high-quality natural images 🖼️.

  • This TSF foundation model is conceptually different from the existing models (text-based 📝 or time series-based 📈), but it shows a comparable or even better performance without any adaptation on time series data.

• We reformulate the TSF task as an image reconstruction task, which is further processed by a visual masked autoencoder (MAE).

🔨 Setup

Our repository is built on Time-Series-Library, MAE, and GluonTS. Please install the dependencies through requirements.txt.

🚀 Quick Start

Please refer to demo.ipynb for a quick start on how to forecast time series using VisionTS, with a clear visualization of the image reconstruction. Our main code is under the visionts directory, which can be imported to your project for use.

📊 Evaluation

Long-Term TSF Benchmarks (Zero-Shot)

We evaluate our methods on 6 long-term TSF benchmarks for zero-shot forecasting. The scripts are under long_term_tsf/scripts/vision_ts_zeroshot. Before running, you should first follow the instructions of Time-Series-Library to download datasets into long_term_tsf/dataset. Using the following command for reproduction:

cd long_term_tsf/
bash scripts/vision_ts_zeroshot/$SOME_DATASET.sh

Monash (Zero-Shot)

We evaluate our methods on 29 Monash TSF benchmarks. Using the following command for reproduction, where the benchmarks will be automatically downloaded.

cd monash/
python run.py

Long-Term TSF Benchmarks (Full-Shot)

We evaluate our methods on 8 long-term TSF benchmarks for full-shot forecasting. The scripts are under long_term_tsf/scripts/vision_ts_fullshot. Using the following command for reproduction:

cd long_term_tsf/
bash scripts/vision_ts_fullshot/$SOME_DATASET.sh

🔗 Citation

@misc{chen2024visionts,
      title={VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasters}, 
      author={Mouxiang Chen and Lefei Shen and Zhuo Li and Xiaoyun Joy Wang and Jianling Sun and Chenghao Liu},
      year={2024},
      eprint={2408.17253},
      archivePrefix={arXiv},
      url={https://arxiv.org/abs/2408.17253}, 
}