Author/Authors :
alweshah, mohammed al-balqa applied university - prince abdullah bin ghazi faculty of communication and information technology - department of computer science, Salt, Jordan , ramadan, enas al-balqa applied university - prince abdullah bin ghazi faculty of communication and information technology - department of computer science, Salt, Jordan , ryalat, mohammad hashem al-balqa applied university - prince abdullah bin ghazi faculty of communication and information technology - department of computer science, Salt, Jordan , almi ani, muder al-hussein bin talal university - computer information systems department, Ma an, Jordan , hammouri, abdelaziz i. al-balqa applied university - prince abdullah bin ghazi faculty of communication and information technology - department of computer science, Salt, Jordan
Abstract :
Classification is a crucial step in data mining as it facilitates decision-making in many areas of human activity, such as scientific endeavors, marketing campaigns, biomedical research and industrial applications. The probabilistic neural network (PNN) is widely utilized to solve classification and pattern recognition problems and is considered an effective method for solving such problems. In this paper, we propose an improved PNN model that employs the water evaporation algorithm (WEA) in order to solve classification problems more efficiently. The proposed method is able to obtain classification accuracies that are close to each other across all 11 benchmark tested datasets from the UCI machine-learning repository, which demonstrates the validity of this method (with respect to classification accuracy). The results show that the WEA is better than the firefly algorithm (FA) and biogeography-based optimization (BBO) in terms of both classification accuracy and convergence speed.
Keywords :
Classification problems , Data mining , Metaheuristics , Probabilistic neural networks , Water evaporation algorithm
