UNIVERSITY OF CALIFORNIA IRVINE Covering Trees: New Variational Bounds for MAP Estimation in Markov Random Fields
نویسندگان
چکیده
ion of Thesis Covering Trees: New Variational Bounds for MAP Estimation in Markov
منابع مشابه
Reasoning and Decisions in Probabilistic Graphical Models - A Unified Framework
OF THE DISSERTATION Reasoning and Decisions in Probabilistic Graphical Models – A Unified Framework By Qiang Liu Doctor of Philosophy in Computer Science University of California, Irvine, 2014 Prof. Alexander Ihler, Chair Probabilistic graphical models such as Markov random fields, Bayesian networks and decision networks (a.k.a. influence diagrams) provide powerful frameworks for representing a...
متن کاملComparison of the bleeding cricothyrotomy model to SimMan for training students and residents emergency cricothyrotomy
Introduction: A cricothyroidotomy is an emergency procedurethat few emergency medicine residents see or perform duringtheir training. Therefore, there is a need for low cost, high fidelitymodels for training. In this study, we explore a new training modelfor cricothyroidotomies (the bleeding CRIC [cost-effective realisticinteractive cricothyroidotomy]) to determine...
متن کاملNew Outer Bounds on the Marginal Polytope
We give a new class of outer bounds on the marginal polytope, and propose a cutting-plane algorithm for efficiently optimizing over these constraints. When combined with a concave upper bound on the entropy, this gives a new variational inference algorithm for probabilistic inference in discrete Markov Random Fields (MRFs). Valid constraints on the marginal polytope are derived through a series...
متن کاملNational Trends in the Use of Inpatient Hospitalization for Combined Abdominoplasty and Breast Surgery
BACKGROUND Combined procedures involving elective breast surgery at the time of abdominoplasty are frequently performed procedures in aesthetic plastic surgery. While found to be safe outpatient procedures, many surgeons elect to perform combined abdominoplasty/breast surgery as inpatient surgery. This study was performed to explore the practice of performing the combined procedure as an inpat...
متن کاملMAP estimation via agreement on (hyper)trees: Message-passing and linear programming
We develop and analyze methods for computing provably optimal maximum a posteriori (MAP) configurations for a subclass of Markov random fields defined on graphs with cycles. By decomposing the original distribution into a convex combination of tree-structured distributions, we obtain an upper bound on the optimal value of the original problem (i.e., the log probability of the MAP assignment) in...
متن کامل