Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Code for our SIGKDD'22 paper Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting.
| Date | Stars |
|---|---|
| 2026-07-31 | 429 |
| 2026-08-06 | 429 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# <div align="center"> Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting </div> <div align="center"> [](https://github.com/zezhishao/BasicTS) [](https://github.com/cnstark/easytorch) [](https://github.com/zezhishao/BasicTS/blob/master/LICENSE) Code for our SIGKDD'22 paper: "[Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting](https://arxiv.org/abs/2206.09113)". The code is developed with [BasicTS](https://github.com/zezhishao/BasicTS), a PyTorch-based benchmark and toolbox for time series forecasting. </div> <img src="figure/STEP.png" alt="TheTable" style="zoom:42%;" /> > Multivariate Time Series (MTS) forecasting plays a vital role in a wide range of applications. Recently, Spatial-Temporal Graph Neural Networks (STGNNs) have become increasingly popular MTS forecasting methods. STGNNs jointly model the spatial and temporal patterns of MTS through graph neural networks and sequential models, significantly improving the prediction accuracy. But limited by model complexity, most STGNNs only consider short-term historical MTS data, such as data over the past one hour. However, the patterns of time series and the dependencies between them (i.e., the temporal and spatial patterns) need to be analyzed based on long-term historical MTS data. To address this issue, we propose a novel framework, in which STGNN is Enhanced by a scalable time series Pre-training model (STEP). Specifically, we design a pre-training model to efficiently learn temporal patterns from very long-term history time series (e.g., the past two weeks) and generate segment-level representations. These representations provide contextual information for short-term time series input to STGNNs and facilitate modeling dependencies between time series. Experiments on three public real-world datasets demonstrate that our framework is capable of significantly enhancing downstream STGNNs, and our pre-training model aptly captures temporal patterns. ## 📚 Table of Contents ```text basicts --> The BasicTS, which provides standard pipelines for training MTS forecasting models. Don't worry if you don't know it, because it doesn't prevent you from understanding STEP's code. datasets --> Raw datasets and preprocessed data figures --> Some figures used in README. scripts --> Data preprocessing scripts. step --> The implementation of STEP, including the architecture, dataloader, loss, and runner for STEP. tsformer_ckpt --> Pre-trained TSFormer for METR-LA, PEMS-BAY, and PEMS04 dataset. training_logs --> Training logs of STEP and TSFormer. ``` ## 💿 Requirements The code is built based on Python 3.9, PyTorch 1.10.0, and [EasyTorch](https://github.com/cnstark/easytorch). You can install PyTorch following the instruction in [PyTorch](https://pytorch.org/get-started/locally/). For example: ```bash pip install torch==1.10.0+cu111 torchvision==0.11.0+cu111 torchaudio==0.10.0 -f https://download.pytorch.org/whl/torch_stable.html ``` After ensuring that PyTorch is installed correctly, you can install other dependencies via: ```bash pip install -r requirements.txt ``` ## 📦 Data Preparation ### **Download Raw Data** You can download all the raw datasets at [Google Drive](https://drive.google.com/file/d/1PY7IZ3SchpyXfNIXs71A2GEV29W5QCv2/view?usp=sharing) or [Baidu Yun](https://pan.baidu.com/s/1CXLxeHxHIMWLy3IKGFUq8g?pwd=blf8), and unzip them to `datasets/raw_data/`. ### **Pre-process Data** You can pre-process all data via: ```bash cd /path/to/your/project bash scripts/data_preparation/all.sh ``` Then the `dataset` directory will look like this: ```text datasets ├─METR-LA ├─METR-BAY ├─PEMS04 ├─raw_data | ├─PEMS04
Excerpt of 8,469 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:5489f32722bd80cc, llm:Repository topics: graph-neural-networks, multivariate-time-series, pre-training, traffic-forecasting; description: 'Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting.'
matched fp:5489f32722bd80cc, llm:Repository topics: graph-neural-networks, multivariate-time-series, pre-training, traffic-forecasting; description: 'Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting.'
matched fp:5489f32722bd80cc, llm:Repository topics: graph-neural-networks, multivariate-time-series, pre-training, traffic-forecasting; description: 'Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting.'