AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Learning Bayesian Network Classifiers to Minimize Class Variable Parameters

Shouta Sugahara, Koya Kato, James Cussens, Maomi Ueno

JMLR 2026 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Approximate inference in Bayesian neural networks

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
minimize class variable parameters (BNOC-BFS) instead of GBN-BDeuBayesian network classifier · 24 UCI datasetsaccuracy: .8111 -> .8385Improved
minimize class variable parameters (BNOC-BFS) instead of ANB-BDeuBayesian network classifier · 24 UCI datasetsaccuracy: .8201 -> .8385No clear effect
minimize NCP instead of all parameters (BNOC-BFS vs GBN-BDeu)Bayesian network classifier · 24 UCI datasetsNCP: 10402.1 -> 1289.7Improved
minimize NCP instead of all parameters on small sample HepatitisBayesian network classifier · Hepatitisaccuracy: .6125 -> .7875Improved
add branch-and-bound depth-first search (DFBnB) instead of breadth-first search (ablation)BNOC · 24 UCI datasetsruntime (s): 1026.0 -> 49.4Improved
add branch-and-bound edge cutting (DFBnB vs BFS) (ablation)BNOC · 24 UCI datasetsaccuracy: .8385 -> .8354No clear effect
use integer programming with NCP penalty (BNOC-IP) instead of BFSBNOC · 24 UCI datasetsaccuracy: .8385 -> .7983Hurt
use integer programming (BNOC-IP) instead of DFBnB on large datasetsBNOC · 12 large UCI datasetsaccuracy: .8880 -> .9075Improved
use BNOC-BFS instead of random forest on sparse dataClassifier · UCI small SPP datasets 1-12accuracy: .8779 -> .8498Hurt
use BNOC-BFS instead of random forest on dense dataClassifier · UCI large SPP datasets 13-24accuracy: .8134 -> .8271Improved
add chordalization and maximum CLL parameter estimation (PCMCE) to BNOC-BFSBNOC-BFS · UCI large SPP datasets 13-24accuracy: .8271 -> .8389Improved
add chordalization and maximum CLL estimation (PCMCE) to BNOC-BFSBNOC-BFS · UCI small SPP datasets 1-12accuracy: .8498 -> .8508No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy