NeurIPS2021

Approximating the Permanent with Deep Rejection Sampling

Juha Harviainen, Antti Röyskö, Mikko Koivisto

5 citations

Abstract

We present a randomized approximation scheme for the permanent of a matrix with nonnegative entries. Our scheme extends a recursive rejection sampling method of Huber and Law (SODA 2008) by replacing the upper bound for the permanent with a linear combination of the subproblem bounds at a moderately large depth of the recursion tree. This method, we call deep rejection sampling, is empirically shown to outperform the basic, depth-zero variant, as well as a related method by Kuck et al. (NeurIPS 2019). We analyze the expected running time of the scheme on random (0,1)(0, 1)-matrices where each entry is independently 11 with probability pp. Our bound is superior to a previous one for pp less than 1/51/5, matching another bound that was known to hold when every row and column has density exactly pp.