https://github.com/ddz16/preformer
This repository contains the pytorch code for the 2023 ICASSP paper "Preformer: Predictive Transformer with Multi-Scale Segment-wise Correlations for Long-Term Time Series Forecasting”
https://github.com/ddz16/preformer
artificial-intelligence correlation-analysis deep-learning deep-neural-networks predictive-modeling time-series time-series-analysis time-series-forecasting time-series-prediction transformer
Last synced: 11 months ago
JSON representation
This repository contains the pytorch code for the 2023 ICASSP paper "Preformer: Predictive Transformer with Multi-Scale Segment-wise Correlations for Long-Term Time Series Forecasting”
- Host: GitHub
- URL: https://github.com/ddz16/preformer
- Owner: ddz16
- License: mit
- Created: 2022-10-26T10:40:01.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2023-02-17T05:51:24.000Z (over 3 years ago)
- Last Synced: 2025-04-10T18:16:37.501Z (about 1 year ago)
- Topics: artificial-intelligence, correlation-analysis, deep-learning, deep-neural-networks, predictive-modeling, time-series, time-series-analysis, time-series-forecasting, time-series-prediction, transformer
- Language: Python
- Homepage: https://arxiv.org/abs/2202.11356
- Size: 522 KB
- Stars: 45
- Watchers: 2
- Forks: 5
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Preformer
This repository contains the pytorch code for the 2023 ICASSP paper "[Preformer: Predictive Transformer with Multi-Scale Segment-wise Correlations for Long-Term Time Series Forecasting](https://arxiv.org/pdf/2202.11356.pdf)”.
# Model
The core MSSC module:
# Acknowledgment
This repository uses some code from [Autoformer](https://github.com/thuml/Autoformer) and [Informer](https://github.com/zhouhaoyi/Informer2020). Thanks to the authors for their work!
# Get Started
## Environment
Install Python 3.6, PyTorch 1.9.0.
## Data
We use all the datasets provided by [Autoformer](https://github.com/thuml/Autoformer) directly, so you can download them from [Google drive](https://drive.google.com/drive/folders/1ZOYpTUa82_jCcxIdTmyr0LXQfvaM9vIy) provided in [Autoformer](https://github.com/thuml/Autoformer). All the datasets are well pre-processed and can be used easily. After downloading, put these dataset files (such as ETTh1.csv) in the `./data/` folder.
# Train
```
python run.py --is_training 1 --root_path ./data/ --data_path ETTh1.csv --model_id ETTh1_96_48 --model Preformer --data ETTh1 --features M --seq_len 96 --label_len 48 --pred_len 48 --e_layers 2 --d_layers 1 --factor 4 --enc_in 7 --dec_in 7 --c_out 7 --des 'Exp' --itr 1 --n_heads 4 --d_model 32 --d_ff 128
```
# Test
```
python run.py --is_training 0 --root_path ./data/ --data_path ETTh1.csv --model_id ETTh1_96_48 --model Preformer --data ETTh1 --features M --seq_len 96 --label_len 48 --pred_len 48 --e_layers 2 --d_layers 1 --factor 4 --enc_in 7 --dec_in 7 --c_out 7 --des 'Exp' --itr 1 --n_heads 4 --d_model 32 --d_ff 128
```
# Main Results
# Citation
```
@inproceedings{du2022preformer,
title={Preformer: Predictive Transformer with Multi-Scale Segment-wise Correlations for Long-Term Time Series Forecasting},
author={Du, Dazhao and Su, Bing and Wei, Zhewei},
booktitle={ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
year={2023},
organization={IEEE}
}
```