GitHub repositories | Zezhi Shao 邵泽志

BasicTS

A fair and scalable time series forecasting benchmark and toolkit for model development, training, and evaluation.

Pre-training on long historical time series to enhance spatial-temporal graph neural networks for multivariate forecasting.

A simple and efficient multivariate time series forecasting baseline built on spatial and temporal identity information.

A decoupled dynamic spatial-temporal graph neural network that models diffusion and inherent signals for traffic forecasting.

A time series pre-training corpus that uses balanced sampling to improve data diversity for universal forecasting models.

A tool for tracking arXiv papers.