English

Concentration of Markov chains with bounded moments

Probability 2019-06-19 v1 Functional Analysis

Abstract

Let {Wt}t=1\{W_t\}_{t=1}^{\infty} be a finite state stationary Markov chain, and suppose that ff is a real-valued function on the state space. If ff is bounded, then Gillman's expander Chernoff bound (1993) provides concentration estimates for the random variable f(W1)++f(Wn)f(W_1)+\cdots+f(W_n) that depend on the spectral gap of the Markov chain and the assumed bound on ff. Here we obtain analogous inequalities assuming only that the qq'th moment of ff is bounded for some q2q \geq 2. Our proof relies on reasoning that differs substantially from the proofs of Gillman's theorem that are available in the literature, and it generalizes to yield dimension-independent bounds for mappings ff that take values in an Lp(μ)L_p(\mu) for some p2p\ge 2, thus answering (even in the Hilbertian special case p=2p=2) a question of Kargin (2007).

Keywords

Cite

@article{arxiv.1906.07260,
  title  = {Concentration of Markov chains with bounded moments},
  author = {Assaf Naor and Shravas Rao and Oded Regev},
  journal= {arXiv preprint arXiv:1906.07260},
  year   = {2019}
}