Fisher’s discriminant modeling for human platelet parameters

Abstract

This study employed multivariate modeling techniques to explore the relationship between platelet parameters and demographic factors. Assumptions such as normality, multicollinearity, and covariance matrix equality were rigorously assessed. Data normalization using the Box-Cox method improved normality, facilitating more robust analyses. Multivariate analysis of Variance revealed significant variations in platelet parameters across demographic groups. Furthermore, linear discriminant analysis demonstrated the capacity to classify individuals based on platelet parameters, particularly concerning gender. The findings underscore the importance of platelet parameters in understanding population characteristics and their potential implications for medical research and clinical practice.