Search Machine Learning Repository:
Taming the Curse of Dimensionality: Discrete Integration by Hashing and Optimization
Authors: Stefano Ermon, Carla Gomes, Ashish Sabharwal and Bart Selman
Conference: Proceedings of the 30th International Conference on Machine Learning (ICML-13)
Abstract: Integration is affected by the curse of dimensionality and quickly becomes intractable as the dimensionality of the problem grows. We propose a randomized algorithm that, with high probability, gives a constant-factor approximation of a general discrete integral defined over an exponentially large set. This algorithm relies on solving only a small number of instances of a discrete combinatorial optimization problem subject to randomly generated parity constraints used as a hash function. As an application, we demonstrate that with a small number of MAP queries we can efficiently approximate the partition function of discrete graphical models, which can in turn be used, for instance, for marginal computation or model selection.
authors venues years
Suggest Changes to this paper.
Brought to you by the WUSTL Machine Learning Group. We have open faculty positions (tenured and tenure-track).