Anytime Anyspace AND/OR Best-First Search for Bounding Marginal MAP

Authors: Qi Lou, Rina Dechter, Alexander Ihler

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical evaluation on three challenging benchmarks demonstrates that our unified best-first search algorithm using pre-compiled variational heuristics often provides tighter anytime upper bounds compared to those state-of-the-art baselines.
Researcher Affiliation Academia Qi Lou University of California, Irvine Irvine, CA 92697, USA qlou@ics.uci.edu Rina Dechter University of California, Irvine Irvine, CA 92697, USA dechter@ics.uci.edu Alexander Ihler University of California, Irvine Irvine, CA 92697, USA ihler@ics.uci.edu
Pseudocode Yes Algorithm 1 Anytime UBFS for MMAP
Open Source Code No The paper does not provide an explicit statement about open-sourcing the code or a link to a code repository.
Open Datasets Yes The benchmark set includes three problem domains: grid networks (grid), medical diagnosis expert systems (promedas), and protein, made from the small protein side-chains of Yanover and Weiss (2002). grid is a subset of the grid dataset used in Marinescu et al. (2017)... promedas is the same dataset as that used in Marinescu et al. (2017).
Dataset Splits No The paper describes experimental settings like using 50% or 10% of variables as MAX variables but does not specify a training, validation, and test split for the datasets themselves.
Hardware Specification No The paper specifies memory allocation ("4GB memory") but does not mention specific hardware components such as CPU or GPU models.
Software Dependencies No The paper mentions that implementations are "in C/C++" but does not specify any software versions for compilers, libraries, or other dependencies.
Experiment Setup Yes The time budget is set to 1 hour for all experiments. We allot 4GB memory to all algorithms, with 1GB extra memory to AAOBF for caching.