Your cheat sheet to the time-series paper deluge

What it does This repository is a curated awesome-list that collects recent time-series research papers and their code from major AI and data-mining venues. The maintainers sort entries into categories like forecasting, anomaly detection, causal discovery, and foundation models, tagging each with its conference source and year. It functions as a living bibliography rather than a software project.

The interesting bit The list is already indexing 2026 papers from KDD, WSDM, ICDE, WWW, and ICLR alongside earlier years, which is unusually current for a community-curated index. The maintainers also flag personally recommended papers with stars and highly-cited work with hearts, giving you a quick heuristic for where to start reading.

Key highlights

• Covers a dozen sub-fields, from irregular time-series learning and early classification to spatio-temporal forecasting and representation learning.
• Each entry links directly to the paper and, when available, to an official code repository.
• Tracks major venues including NeurIPS, ICML, KDD, AAAI, ICLR, WWW, and ICDE with frequent batch updates.
• Uses emoji markers to highlight notable or highly-cited papers at a glance.

Caveats

• This is strictly a reading list; there is no installable framework or unified toolkit underneath.
• The star and heart ratings reflect the maintainers’ personal curation rather than a formal peer-review or automated citation-ranking process.

Verdict Use this if you need to scan the latest time-series literature without opening ten separate conference proceedings. Look elsewhere if you want a drop-in Python library or reproducible benchmark suite.