Extracting Information in Agricultural Data Using Fuzzy-Rough Sets Hybridization and Clonal Selection Theory Inspired Algorithms

Abstract

Mining agricultural data with artificial immune system (AIS) algorithms, particularly the clonal selection algorithm (CLONALG) and artificial immune recognition system (AIRS), form the bedrock of this paper. The fuzzy-rough feature selection (FRFS) and vaguely quantified rough set (VQRS) feature selection are coupled with CLONALG and AIRS for improved detection and computational efficiencies. Comparative simulations with sequential minimal optimization and multi-layer perceptron reveal that the CLONALG and AIRS produced significant results. Their respective FRFS and VQRS upgrades namely, FRFS-CLONALG, FRFS-AIRS, VQRS-CLONALG, and VQRS-AIRS, are able to generate the highest detection rates and lowest false alarm rates. Thus, gathering useful information with the AIS models can help to enhance productivity related to agriculture.