Tsallis-regularized optimal transport unifies Monge and Sinkhorn frameworks (AAAI)
FrnkNlsn · x · 2026-10-16
FrnkNlsn shares his AAAI paper Tsallis Regularized Optimal Transport and Ecological Inference (arXiv:1609.04495). Optimal transport is computationally costly; entropy regularization with Sinkhorn balancing speeds it up. This work replaces Shannon entropy with Tsallis entropy, unifying Monge-Kantorovitch and Sinkhorn-Cuturi OT into one framework that interpolates Wasserstein through KL divergence (including Pearson, Neyman, Hellinger), generalizes known metric properties, and offers efficient algorithms with convergence proofs. It also pioneers OT for ecological inference — reconstructing joint distributions from marginals — demonstrated on 2012 US presidential election data linking ethnic groups and voter preferences.
More from Research
- OneWM: unified causal training shows world and action modeling improve each other — IAAI233 · 2026-10-16
- Baidu's Virtual Head Attention tames the KV cache, FLOPs, quality triangle in long-context LLMs — baidu · 2026-10-16
- 5.8M-param Speech Event Detector adds laughs, breaths and背景 sounds to Whisper transcripts — Ayush Kumar Bar · 2026-10-16
- Huawei's SUPRA-Tex turns one image into relightable 3D materials with WYSIWYG fidelity — Chen Yang · 2026-10-16
- RoboVista benchmark: Gemini-4 Argon hits 81.9% on robotics understanding — berkeley_ai · 2026-10-16
- DeepMind details how its AI models help detect, treat and prevent disease — GoogleDeepMind · 2026-10-16