Mining Drug Data via Mixtures and Corresponding Computational Issues

• Activity Number: 479
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Learning and Data Mining
• Author(s): Xu (Sunny) Wang*+ and Hugh A. Chipman
• Companies: St. Francis Xavier University and Acadia University
• Keywords: Mixture Normals ; Disriminant Analysis ; EM Algorithm ; Penalized MLE

Abstract

We proposed and developed constrained mixture discriminant analysis (CMDA), a model-based statistical learning method. The main idea of CMDA is to model the distribution of the observations given the class label (e.g. active or inactive class) as a constrained mixture distribution, and then use Bayes' rule to predict the probability of being active for each observation in the testing set. Constraints are used to deal with the otherwise explosive growth of the number of parameters with increasing dimensionality. CMDA is designed to solve several challenges in modeling drug data sets, such as multiple mechanisms (which correspond to mixture components), the rare target problem (i.e. imbalanced classes), and the identification of relevant subspaces of descriptors (i.e. variable selection).