Contact Us Search Paper

2019 Scientific Conference on Network, Power Systems and Computing , Pages 132-135

A classification method based on linear discriminant analysis and multivariate adaptive spline

Zhihui Li, Jiaxin Liu, Yufei, Huang, Jia Xu

Corresponding Author:

Zhihui Li

Abstract:
In this paper, a data classification method based on linear discriminant analysis and multivariate adaptive splines is proposed, and this method is a combination of dimension reduction and classification. Firstly, the most effective classification features are determined by linear discriminant analysis. Then, the input variables are divided into intervals by multiple adaptive regression splines (MARS), and the non-linear classification is transformed into the linear classification problem. Finally, classification is realized by perceptron. Experiment results prove that the method is effective both in classification performance and predicting speed.
Keywords:
Machine learning, data dimension reduction, classification, multiple adaptive regression splines
Cite this paper:
Zhihui Li, Jiaxin Liu, Yufei, Huang, Jia Xu, A classification method based on linear discriminant analysis and multivariate adaptive spline. 2019 Scientific Conference on Network, Power Systems and Computing (NPSC 2019), 2019: 132-135. DOI: https://doi.org/10.33969/EECS.V3.030.