MDL-GBC constructs class-conditional granular balls by comparing single-ball, two-ball, and core-boundary models under a unified MDL criterion and aggregates them for prediction, achieving the best average accuracy and Macro-F1 on 18 benchmarks.
An efficient spectral clustering algorithm based on granular-ball
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
MDL-GBG selects the shortest-description-length model among single-ball, two-ball, and core-ball-plus-residual options to generate stable granular balls that improve downstream clustering on UCI datasets.
citing papers explorer
-
A Boundary-Aware Non-parametric Granular-Ball Classifier Based on Minimum Description Length
MDL-GBC constructs class-conditional granular balls by comparing single-ball, two-ball, and core-boundary models under a unified MDL criterion and aggregates them for prediction, achieving the best average accuracy and Macro-F1 on 18 benchmarks.
-
MDL-GBG: A Non-parametric and Interpretable Granular-Ball Generation Method for Clustering
MDL-GBG selects the shortest-description-length model among single-ball, two-ball, and core-ball-plus-residual options to generate stable granular balls that improve downstream clustering on UCI datasets.