A new discrete particle swarm algorithm applied to attribute selection in a bioinformatics data set

摘要: Many data mining applications involve the task of building a model for predictive classification. The goal such is to classify examples (records or instances) into classes categories same type. use variables (attributes) not related can reduce accuracy and reliability classification prediction model. Superuous also increase costs - particularly on large sets. We propose discrete Particle Swarm Optimization (PSO) algorithm designed attribute selection. proposed deals with variables, its population candidate solutions contains particles different sizes. performance this compared standard binary PSO selecting attributes in bioinformatics set. criteria used comparison are: (1) maximizing accuracy; (2) finding smallest subset attributes.