We examine two analytical characterisation of the metastable behavior of a Markov chain. The first one expressed in terms of its transition probabilities, and the second one in terms of its large deviations rate functional. Consider a sequence of continuous-time Markov chains $(X^{ (n)}_t:t\ge 0)$ evolving on a fixed finite state space $V$. Under a hypothesis on the jump rates, we prove the existence of times-scales $\theta^{ (p)}_n$ and probability measures with disjoint supports $\pi^{ (p)}_j$, $j\in S_p$, $1\le p \le q$, such that (a) $\theta^{ (1)}_n \to \infty$, $\theta^{ (k+1)}_n/\theta^{ (k)}_n \to \infty$, (b) for all $p$, $x\in V$, $t>0$, starting from $x$, the distribution of $X^{ (n)}_{ t \theta^{ (p)}_n}$ converges, as $n\to\infty$, to a convex combination of the probability measures $\pi^{ (p)}_j$. The weights of the convex combination naturally depend on $x$ and $t$. Let $I_n$ be the level two large deviations rate functional for $X^{ (n)}_t$, as $t\to\infty$. Under the same hypothesis on the jump rates and assuming, furthermore, that the process is reversible, we prove that $I_n$ can be written as $I_n = I^{ (0)} \,+\, \sum_{ 1\le p\le q} (1/\theta^{ (p)}_n) \, I^{ (p)}$ for some rate functionals $I^{ (p)}$ which take finite values only at convex combinations of the measures $\pi^{ (p)}_j$: $I^{ (p)}(\mu) < \infty$ if, and only if, $\mu = \sum_{ j\in S_p} \omega_j\, \pi^{ (p)}_j$ for some probability measure $\omega$ in $S_p$.

Categories: math.PR cond-mat.stat-mech

Metastable $\Gamma$-expansion of finite state Markov chains level two large deviations rate functions