An offline data-driven process for learning operator selection from metaheuristic search traces

Download PDF
Open Webpage

Panagiotis Kalatzantonakis, Angelo Sifaleras, Nikolaos Samaras

Published: 2025, Last Modified: 11 Dec 2025Swarm Evol. Comput. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•TRACE learns decision policies from features extracted during metaheuristic execution.•TRACE-VNS improves convergence and solution quality over conventional GVNS.•Offline-trained models guide operator selection with zero runtime tuning.•Models generalize across 84 CVRP instances with strong PR-AUC performance.•Learns to adapt across problems by modeling search dynamics, not problem rules.