Minimum Spanning Trees ensemble framework: A geometric classifier

doi:10.1016/j.knosys.2026.117057

Minimum Spanning Trees ensemble framework: A geometric classifier

2026-09-22

0

OA

AI

V

Víctor Aceña *

F

F. Javier Martín-Campo

P

Paula Terán-Viadero

DOI：10.1016/j.knosys.2026.117057

Original

Original request for help

Save

Abstract

En 中文

• An ensemble of MSTs resampled across diverse class ratios improves robustness. • Introduces the affinity score, a continuous measure of class membership strength. • A two-phase design separates structure learning from prediction, avoiding graph reconstruction. • The pre-computed score supports interchangeable prediction strategies, like k-NN.

Keywords:

Minimum Spanning Tree (MST)

Ensemble learning

Affinity score

Inductive learning

Supervised classification

Minimum Spanning Trees ensemble framework: A geometric classifier

Abstract

Journal

Organization

Cited Papers