Multiclass Linear Discriminant Analysis with Ultrahigh-Dimensional Features

Yi Li, Biostatistics Department, University of Michigan, USA

Within the framework of Fisher’s discriminant analysis, we propose a multiclass classification method, which embeds variable screening for ultrahigh-dimensional predictors. Leveraging inter-feature correlations, we show that the proposed linear classifier recovers informative features with probability tending to one and can asymptotically achieve a zero misclassification rate. We evaluate the finite sample performance of the method via extensive simulations and use this method to classify post-transplantation rejection types based on patients’ gene expressions.

Please direct all questions to the following contacts - Your question will be routed to get you an answer as quickly as possible.

• MSUStatSymposium2018@stt.msu.edu
• Michigan State University
  Department of Statistics and Probability
  Room C413 Wells Hall
  619 Red Cedar Road
  East Lansing, MI 48824