Low-variance Black-box Gradient Estimates for the Plackett-Luce Distribution

Autor:	Gadetsky, Artyom, Struminsky, Kirill, Robinson, Christopher, Quadrianto, Novi, Vetrov, Dmitry
Rok vydání:	2019
Předmět:	Computer Science - Machine Learning Statistics - Machine Learning
Druh dokumentu:	Working Paper
Popis:	Learning models with discrete latent variables using stochastic gradient descent remains a challenge due to the high variance of gradient estimates. Modern variance reduction techniques mostly consider categorical distributions and have limited applicability when the number of possible outcomes becomes large. In this work, we consider models with latent permutations and propose control variates for the Plackett-Luce distribution. In particular, the control variates allow us to optimize black-box functions over permutations using stochastic gradient descent. To illustrate the approach, we consider a variety of causal structure learning tasks for continuous and discrete data. We show that our method outperforms competitive relaxation-based optimization methods and is also applicable to non-differentiable score functions. Comment: Accepted as a conference paper at AAAI 2020. Shortened version of the paper appears at BDL NeurIPS 2019 workshop
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/1911.10036 Zobrazit plný text záznamu View this record from Arxiv