Tracking the flood of time-series papers so you don't have to

What it does

This repository is a curated bibliography that catalogs time-series papers accepted at major CS conferences—NeurIPS, ICML, KDD, AAAI, WWW, and others—and sorts them into task-specific buckets such as multivariate forecasting, anomaly detection, imputation, demand prediction, and traffic location prediction. Each entry typically links to the paper and to code when the authors have released it. The maintainer, a UNSW Sydney PhD student, organizes the list through markdown tables with custom methodology abbreviations.

The interesting bit

Instead of treating GitHub as a code host, the project treats it as a living research database. The granularity is the selling point: papers are tagged by both task and methodology (with an in-repo abbreviation scheme covering everything from CDEs to Transformers), making it possible to scan what, say, spatio-temporal GAN work looked like across venues last year. It also aggressively tracks the recent wave of time-series foundation models and LLM-adjacent papers.

Key highlights

• Covers 10+ top-tier venues including ICML, NeurIPS, ICLR, KDD, AAAI, and ICDE
• Task-oriented sections: forecasting, imputation, anomaly detection, generation, travel-time estimation, stock prediction, and more
• Links to both papers and official code repositories when available
• Tracks emerging directions such as time-series foundation models and medical LLM reasoning
• Maintained by an active researcher in the field

Caveats

• This is a curated index, not a software library; there is no installable code or framework here
• Code availability is inconsistent—many recent entries list no associated repository
• The README is extensive and still growing; navigation relies on manual scrolling through markdown tables

Verdict

Time-series researchers who need a quick, broad survey of what got published where will find this saves hours of digging. If you are hunting for a reusable training pipeline or benchmark suite, this is not that.