Deep Learning in Science and Engineering
- Texas A&M University
- College Station, TX
- Joe C. Richardson Petroleum Engineering Building (RICH) 910
- Ulisses Braga-Neto, Department of Electrical & Computer Engineering
- Pattern Recognition for Small-Sample Applications
Abstract
Modern applications often produce large amounts of data characterized by a very large number of measurements made on a much smaller number of sample points. This introduces difficult challenges in the application of classification methods to obtain accurate predictive models. In this talk, we discuss our recent work on this topic, highlighting Bayesian approaches to classification, as well as novel approaches for classification and error estimation based on restricted and nonstationary data.