A framework of multivariant statistical model based tool using particle swarm optimization with fuzzy data for the classification of yeast data

Yeast is one of the major components for the formation of different medicines and various chemical products. So, yeast data classification is one of the major bioinformatics task. If the type of yeast can be categorized at primary stages based on initial characteristics of it, a lot of technical procedure can be avoided in the preparation of chemical and medical products. In this paper, an effort has been taken to classify yeast data. The yeast dataset, obtained from UCI machine learning laboratories, is used. Here, 50 selected data samples have been chosen for case study. At first Total effect of each selected samples has been calculated with Factor analysis (FA) and principal component analysis (PCA). On the basis of average error of FA and PCA, Total Effect obtained from factor analysis has been chosen for applying two soft computing models, they are Fuzzy Time Series model (FTS), and Particle Swarm Optimization model respectively. The performance of these two soft computing models is then evaluated using residual analysis.