AIISR - AI INSPIRED STOPPING RULE FOR GRASP METAHEURISTIC

Vol 53, 2021 - 139666
Trabalho completo (oral)
Favorite this paper
How to cite this paper?
Abstract

Metaheuristics are high level heuristics and usually lack an effective stopping criteria, relying on generic information, like number of iterations or time, to decide when they should stop. To tackle this problem for the GRASP metaheuristic, past works proposed a probabilistic stopping criteria based on the fact that the solutions of GRASP iterations follow a normal distribution. However, that method has two drawbacks: it is slow to estimate the probabilities and it is not very accurate. In order to address those problems, the present work proposes an alternative method that makes use of Decision Tree Regression to produce a stopping rule. In this paper, we show that this approach can be faster and more precise. We demonstrate that, in some cases, models trained on an instance of a given problem can also be applied to instances belonging to the same problem or even to other problems.

Share your ideas or questions with the authors!

Did you know that the greatest stimulus in scientific and cultural development is curiosity? Leave your questions or suggestions to the author!

Sign in to interact

Have a question or suggestion? Share your feedback with the authors!

Institutions
  • 1 Universidade Federal do Rio de Janeiro
  • 2 Federal University of Rio de Janeiro
Track
  • 12 - CI - Computational Intelligence
Keywords
Stopping Rule
grasp
Machine Learning Aided Optimization