multivariate-time-series

This repository contains a reading list of papers on Time Series Segmentation. This repository is still being continuously improved.

Code for our SIGKDD'22 paper Pre-training-Enhanced Spatial-Temporal Graph Neural Network For Multivariate Time Series Forecasting.

Code for our CIKM'22 paper Spatial-Temporal Identity: A Simple yet Effective Baseline for Multivariate Time Series Forecasting.

Book and material for the course "Time series analysis with Python" (STA-2003)

This repository contains a reading list of papers on multivariate time series anomaly detection. This repository is still being continuously improved.

Unified Python/CLI time-series decomposition library with native acceleration, multivariate workflows, and machine-readable outputs.

📦 A Python package for online changepoint detection, implementing state-of-the-art algorithms and a novel approach based on neural networks.

Finite-Interval Forecasting Engine: Machine learning models for discrete-time survival analysis and multivariate time series forecasting

Inferencing 'PatchTST' and 'Informer' to harness the power of transformers for multivariate 'long sequence time-series forecasting' (LSTF).

[NeurIPS 2025] NeurIPT: Foundation Model for Neural Interfaces

bayesian-sgdlm is a Python script for fully Bayesian SGDLMs, treating each node as a VAR( 𝑝) DLM. It leverages decouple–recouple filtering with Variational Bayes and importance sampling to estimate sparse, time-varying cross-lag dependencies (including pandemic dummies) without ever inverting the full multivariate system.

List of books on time series analysis, with links to code where available

Standard Bayesian VAR with conjugate priors and Minnesota dummy-observation priors (unit-root and cointegration dummies) for analyzing shock transmission between U.S. 5-year and 3-year T-Bills and Colombian 5-year TES.

[AAAI-23] AnoViz: A Visual Inspection Tool of Anomalies in Multivariate Time Series

UQGNN: Uncertainty Quantification of Graph Neural Networks for Multivariate Spatiotemporal Prediction (ACM SIGSPATIAL 2025)

Paper: Multi-Scale Ensemble Booster for Improving Existing TSD Classifiers. We proposed a highly easy-to-use performance enhancement framework called multi-scale ensemble booster(MEB), helping existing time series classification methods achieve performance leap. Our proposed MEB-ResNet achieved the most advanced time series classification ability.

Official repository for the paper "Time Series Kernels based on Nonlinear Vector AutoRegressive Delay Embeddings" (NeurIPS 2023)

Official repository of the RideBench paper: "RideBench: A Large-Scale Exogenous-Aware Benchmark for Ride-Hailing Time Series Forecasting". Benchmarking exogenous-aware forecasting on 4 years of half-hourly data across 200 spatial areas.