Good Parameters for PSO in Optimizing Laying Hen Diet

Abstract

Manual formulation of poultry diet by taking into account the fulfillment of all nutrients requirement with least cost is a difficult task. Particle Swarm Optimization (PSO) shows promising technique to solve this problem. However, there is a lack of studying a good parameter for PSO to solve feed formulation problem since PSO is sensitive to control parameter which depends on the problem. Therefore, this study investigates good swarm size, total iterations, acceleration coefficients, and inertia weight to produce a better formula. PSO with proposed good parameters is compared with other parameters. The obtained result shows that PSO with good parameters choice produces the highest fitness. Furthermore, good parameters of PSO can be used as a reference for a software developer and for further research to optimize poultry diet using PSO.

Effective predictive modelling for coronary artery diseases using support vector machine

“…Effective parameter initialization is situation-dependent, and each situation may require unique parameters. By specifying the appropriate parameters, the model can be guided to its optimal solution [31]. Parameter optimization is time consuming if done manually, especially since it has many parameters.…”