Active Inference · Paper · 2024

Enhancing Population-based Search with Active Inference

ArXiv

Catalog Row34
Citation KeyFriedman2024EnhancingPopulationBasedSearch034
Paper FolderAvailable
DOI10.48550/arXiv.2408.09548

Overview

Extracted from the local paper documentation when available.

This paper proposes the integration of Active Inference into traditional population-based metaheuristics to enhance their performance through anticipatory environmental adaptation. Specifically, it demonstrates this approach using Ant Colony Optimization on the Travelling Salesman Problem, showing improved solutions with minimal computational cost increase.

population searchActive InferenceAnt Colony OptimizationTSPmetaheuristicsanticipatory adaptationcomputational optimization

Use Notes

Concise findings and methods pulled from README/SKILL documentation.

Findings / Concepts
  • Integration of Active Inference into population-based metaheuristics.
  • Application of the proposed method to Ant Colony Optimization for the Travelling Salesman Problem.
  • Demonstration of improved solution quality with marginal computational cost increase.
  • Identification of performance patterns related to graph node number and topology.
Methods / Techniques
  • Utilization of Active Inference framework for anticipatory adaptation.
  • Experimental evaluation of Ant Colony Optimization with Active Inference.
  • Analysis of performance based on varying graph structures.
  • Comparison of traditional reactive approaches with the proposed anticipatory model.

Citation

Plain-text citation for quick reuse.

Friedman, Daniel Ari. 2024. Enhancing Population-based Search with Active Inference. ArXiv.

Primary source Documentation BibTeX