Expectation Algorithm (ExA): A Socio-inspired Optimization Methodology
This paper introduces a new socio-inspired algorithm referred to as Expectation Algorithm (ExA), which is mainly inspired from the society individuals. The ExA modelled the variables of the problems as individuals of a society. The variables select their
- PDF / 317,535 Bytes
- 22 Pages / 439.37 x 666.142 pts Page_size
- 97 Downloads / 205 Views
Abstract This paper introduces a new socio-inspired algorithm referred to as Expectation Algorithm (ExA), which is mainly inspired from the society individuals. The ExA modelled the variables of the problems as individuals of a society. The variables select their values by expecting the values of the other variables minimizing the objective function. The performance of the algorithm is validated by solving 50 unconstrained test problems with dimensions up to 30. The solutions were compared with several recent algorithms such as Covariance Matrix Adaptation Evolution Strategy, Artificial Bee Colony, Comprehensive Learning Particle Swarm Optimization, Selfadaptive Differential Evolution Algorithm, Backtracking Search Optimization Algorithm, Ideology Algorithm and Multi-Cohort Intelligence algorithm. The Wilcoxon signed-rank test was carried out for the statistical analysis and verification of the performance. The results from this study highlighted that the ExA outperformed most of the other algorithms in terms of function evaluations and computational time. The prominent features of the ExA algorithm along with the limitations are discussed as well.
A. S. Shastri (B) · A. Jagetia · A. Sehgal · M. Patel · A. J. Kulkarni Symbiosis Institute of Technology, Symbiosis International (Deemed University), Pune 412115, MH, India e-mail: [email protected] A. Jagetia e-mail: [email protected] A. Sehgal e-mail: [email protected] M. Patel e-mail: [email protected] A. J. Kulkarni e-mail: [email protected]; [email protected] A. J. Kulkarni Odette School of Business, University of Windsor, 401 Sunset Avenue, Windsor, ON N9B3P4, Canada © Springer Nature Singapore Pte Ltd. 2019 A. J. Kulkarni et al. (eds.), Socio-cultural Inspired Metaheuristics, Studies in Computational Intelligence 828, https://doi.org/10.1007/978-981-13-6569-0_10
193
194
A. S. Shastri et al.
Keywords Expectation algorithm · Unconstrained optimization · Socio-inspired optimization method
1 Introduction In recent few years, several metaheuristic algorithms have been proposed from nature-inspired domain, evolutionary-based approaches and physics and biologically-based approaches. The notable nature-inspired methods include algorithms such as Ant Colony Optimization (ACO) [6, 27], Particle Swarm Optimization (PSO) [7, 10], Artificial Bee Colony Algorithm (ABC) [12, 11], Bacterial Foraging Optimization Algorithm (BFO) [5], Bat Algorithm (BA) [35, 36], Cuckoo Search Algorithm (CS) [37], Glowworm Swarm Optimization (GSO) [18], Firefly Algorithm [34], Predator–Prey Algorithm [31], etc. The evolutionary-based algorithms are Genetic Algorithms (GA) [19], Genetic Programming [16], Biogeography-based optimization Algorithm [29], Differential Evolution (DE) [21], Artificial Immune System [32], Memetic Computing Algorithms [2], etc. The Physics and Biologicallybased Methods developed are Black Hole Algorithm [8], Gravitational Search Algorithm [27], River Formation Dynamics [23], Simulated Annealing (SA) [20, 2
Data Loading...