Department of Applied Mathematics & Physics, Kyoto University
Technical Report 2008-004 (April 21, 2008)
Tabu Programming Method: A New Meta-Heuristics Algorithm Using Tree Data Structures for Problem Solving
by Abdel-Rahman Hedar, Emad Mabrouk, and Masao Fukushima
The core of artificial intelligence and machine learning is to get computers to solve problems automatically. One of the great tools that attempt to achieve that goal is Genetic Programming (GP). GP is a generalization procedure of the well-known meta-heuristic of Genetic Algorithms (GAs). Meta-heuristics have shown successful performance in solving many combinatorial search problems. In this paper, we introduce a more general framework of meta-heuristics called Meta-Heuristics Programming (MHP) as general machine learning tools. As an alternative to GP, Tabu Programming (TP) is proposed as a special procedure of MHP frameworks. One of the main features of MHP is to exploit local search in order to overcome some drawbacks of GP, especially high disruption of its main operations; crossover and mutation. We show the efficiency of the proposed TP method through numerical experiments.