Estimation-based Metaheuristics: A New Branch of Computational Intelligence
In this paper, a new branch of computational intelligence named estimation-based metaheuristic is introduced. Metaheuristic algorithms can be classified based on their source of inspiration. Besides biology, physics and chemistry, state estimation algorithm also has become a source of inspiration fo...
| Main Authors: | , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/14583/ http://umpir.ump.edu.my/id/eprint/14583/1/P064%20pg469-476.pdf |
| Summary: | In this paper, a new branch of computational intelligence named estimation-based metaheuristic is introduced. Metaheuristic algorithms can be classified based on their source of inspiration. Besides biology, physics and chemistry, state estimation algorithm also has become a source of inspiration for developing metaheuristic algorithms. Inspired by the estimation capability of Kalman Filter, Simulated Kalman Filter, SKF, uses a population of agents to make estimations of the optimum. Each agent in SKF acts as a Kalman Filter. By adapting the standard Kalman Filter framework, each individual agent finds an optimization solution by using a simulated measurement process that is guided by a best-so-far solution as a reference. Heuristic Kalman Algorithm (HKA) also is inspired by the Kalman Filter framework. HKA however, explicitly consider the optimization problem as a measurement process in generating the estimate of the optimum. In evaluating the performance of the estimation-based algorithms, it is implemented to 30 benchmark functions of the CEC 2014 benchmark suite. Statistical analysis is then carried out to rank the estimation-based algorithms’ results to those obtained by other metaheuristic algorithms. The experimental results show that the estimation-based metaheuristic is a promising approach to solving global optimization problem and demonstrates a competitive performance to some well-known metaheuristic algorithms |
|---|