تحليل مميزي غير خطي انعطاف پذير با استفاده از رگرسيون ناپارامتري

Discriminant analysis is a way for classification of one object or a group to one or more separate groups that are known or unknown. In scientific researches we often use linear or quadratic functions for classification. But in this paper, we suggest a nonlinear discrimination method that uses two nonparametric regression methods, namely multivariate adaptive regression splines and adaptive additive model. In a simulation study, we investigate the application way of proposed methods and comparing them with the ordinary nonlinear discrimination methods via their means of error rates.