Simple Heuristic
Simple heuristics are rules of thumb used to efficiently solve complex problems, particularly in optimization and search, where finding optimal solutions is computationally intractable. Current research focuses on improving heuristic performance through machine learning techniques, such as integrating reinforcement learning agents or graph neural networks to learn effective search strategies and adapt to specific problem instances. This work aims to enhance the speed and solution quality of existing algorithms across diverse applications, from production scheduling and route optimization to AI planning and puzzle solving, bridging the gap between handcrafted heuristics and data-driven approaches.
Papers
High-level hybridization of heuristics and metaheuristics to solve symmetric TSP: a comparative study
Carlos Alberto da Silva Junior, Roberto Yuji Tanaka, Luiz Carlos Farias da Silva, Angelo Passaro
Arithmetic Without Algorithms: Language Models Solve Math With a Bag of Heuristics
Yaniv Nikankin, Anja Reusch, Aaron Mueller, Yonatan Belinkov
A Benchmark for Maximum Cut: Towards Standardization of the Evaluation of Learned Heuristics for Combinatorial Optimization
Ankur Nath, Alan Kuhnle
Deep Symbolic Optimization for Combinatorial Optimization: Accelerating Node Selection by Discovering Potential Heuristics
Hongyu Liu, Haoyang Liu, Yufei Kuang, Jie Wang, Bin Li