On the $\alpha$-lazy version of Markov chains in estimation and testing problems

Open in new window