FTPL Algorithm Gets Sub‑Linear Regret for Online Non‑Convex Opt

FTPL algorithm achieves sub-linear regret for online non-convex optimization with long-term constraints, outperforming prior methods on river-pollutant source identification. Read more: getnews.me/ftpl-algorithm-gets-sub-... #onlineoptimization #ftpl

Distributed Associative Memory Using Online Convex Optimization

Preprint (arXiv:2509.22321) posted 26 Sep 2025 presents an associative memory using gradient descent over routing‑tree links, achieving sub‑linear regret and faster recall. getnews.me/distributed-associative-... #associativememory #onlineoptimization