Ant Colony Sampling with GFlowNets for Combinatorial Optimization

2024/03/11 by Minsu Kim, Kim, Minsu, Sanghyeok Choi +9 · 10 citations
Computer Science · Engineering · #BIM and Construction Integration #FOS: Computer and information sciences #Machine Learning (cs.LG) #Metaheuristic Optimization Algorithms Research #Neural and Evolutionary Computing (cs.NE) #Optimization and Packing Problems

paper · pdf · doi:10.48550/arxiv.2403.07041

openalex publication_date 2024/03/11 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We present the Generative Flow Ant Colony Sampler (GFACS), a novel meta-heuristic method that hierarchically combines amortized inference and parallel stochastic search. Our method first leverages Generative Flow Networks (GFlowNets) to amortize a multi-modal prior distribution over combinatorial solution space that encompasses both high-reward and diversified solutions. This prior is iteratively updated via parallel stochastic search in the spirit of Ant Colony Optimization (ACO), leading to the posterior distribution that generates near-optimal solutions. Extensive experiments across seven combinatorial optimization problems demonstrate GFACS's promising performances.

Cited by

• HeatACO: A Heatmap-Guided Max--Min Ant System for Large-Scale Travelling Salesman Problems
• CogMCTS: A Novel Cognitive-Guided Monte Carlo Tree Search Framework for Iterative Heuristic Evolution with Large Language Models
• Reinforced sequential Monte Carlo for amortised sampling
• VRPAgent: LLM-Driven Discovery of Heuristic Operators for Vehicle Routing Problems
• Hybrid-Balance GFlowNet for Solving Vehicle Routing Problems
• Test-Time Search in Neural Graph Coarsening Procedures for the Capacitated Vehicle Routing Problem
• NeuFACO: Neural Focused Ant Colony Optimization for Traveling Salesman Problem
• Mitigating Hallucinations in LM-Based TTS Models via Distribution Alignment Using GFlowNets
• Hybrid Node-Destroyer Model with Large Neighborhood Search for Solving the Capacitated Vehicle Routing Problem
• Edge-Selector Model Applied for Local Search Neighborhood for Solving Vehicle Routing Problems

Related

• Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNets
• MG2FlowNet: Accelerating High-Reward Sample Generation via Enhanced MCTS and Greediness Control
• Improving GFlowNets with Monte Carlo Tree Search
• Boosting Ant Colony Optimization via Solution Prediction and Machine Learning
• DeepACO: Neural-enhanced Ant Systems for Combinatorial Optimization