Deploying Fuzzy Rough Set and Artificial Immune System Algorithms for Air Quality Prediction

Abstract

The widespread of air pollution due to emissions of highly detrimental concentrates on human life needs paramount attention. Devising air prediction strategy that accurately analyze air quality level through gathering of useful information allow for relevant organizations to disseminate promptly control measures. This paper proposes the use of artificial immune system (AIS) algorithms consisting of Immunos algorithms (Immunos-1, Immunos-2, Immunos-99), CLONal selection ALGorithm (CLONALG), clonal selection classification algorithm (CSCA), artificial immune recognition system algorithms (AIRS1, AIRS2, Parallel AIRS2) for air quality prediction. The fuzzy rough set selects pertinent data features summarizing interpretations of the data. Comparative simulations reveal that the Parallel AIRS2 produced superlative results to other algorithms with detection rate of 96.40%. Effective prediction performance can be generated with AIS algorithms having highest detection rates and lowest false alarm rates.