{"id":"arxiv/2609.21985","url":"https://arxiv.org/abs/2609.21985","tags":["stat.ME","math.ST"],"score":0,"title":"Consistency of Optimal Matching-based Clustering for Mixtures of Markov Chains","domain":"stat.ME","source":"arxiv","authors":["Ottavio Khalifa"],"reviews":0,"subject":"stat.ME","summary":"We study clustering of categorical sequences using the Optimal Matching (OM) distance under finite mixtures of finite-state Markov chains. We show that the normalized OM distance between two independent chains converges almost surely to a deterministic population quantity, concentrates exponentially around its finite-horizon mean, and admits an $O(\\sqrt{\\log n/n})$ convergence rate when the two chains have the same transition kernel. These population quantities yield a natural separation condition: the largest within-component limit must be smaller than the smallest between-component limit.…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-09-18","reactions":{}}