Decomposition Bounds for Marginal MAP

Authors: Wei Ping, Qiang Liu, Alexander T. Ihler

NeurIPS 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We report experimental results in Section 6 and conclude the paper in Section 7. ... In this section, we demonstrate our algorithm on a set of real-world graphical models from recent UAI inference challenges... Figure 2 and Figure 3 compare the convergence of the different algorithms...
Researcher Affiliation Academia Computer Science, UC Irvine Computer Science, Dartmouth College {wping,ihler}@ics.uci.edu qliu@cs.dartmouth.edu
Pseudocode Yes Algorithm 1 Generalized Dual-decomposition (GDD)
Open Source Code No No statement regarding the release of the paper's source code is found.
Open Datasets Yes In this section, we demonstrate our algorithm on a set of real-world graphical models from recent UAI inference challenges, including two diagnostic Bayesian networks with 203 and 359 variables and max domain sizes 7 and 6, respectively, and several MRFs for pedigree analysis with up to 1289 variables, max domain size of 7 and clique size 5.7 We construct marginal MAP problems on these models by randomly selecting half of the variables to be max nodes, and the rest as sum nodes. ... 7See http://graphmod.ics.uci.edu/uai08/Evaluation/Report/Benchmarks.
Dataset Splits No The paper uses real-world graphical models from UAI inference challenges but does not specify the training/validation/test splits, only that half of the variables are randomly selected as max nodes and the rest as sum nodes.
Hardware Specification No No specific hardware details (e.g., GPU/CPU models, memory) used for running the experiments are mentioned in the paper.
Software Dependencies No The paper mentions implementing algorithms and using an 'off-the-shelf L-BFGS implementation', but does not provide specific version numbers for any software dependencies.
Experiment Setup Yes We implement several algorithms that optimize the same primal marginal MAP bound, including our GDD (Algorithm 1), the WMB algorithm in [16] with ibound = 1... We tune the damping rate of WMB from 0.01 to 0.05... The elimination order that we use is obtained by a weighted-min-fill heuristic [1] constrained to eliminate the sum nodes first... In our implementation, we find that a few gradient steps (e.g., 5) with a backtracking line search using the Armijo rule works well in practice.