III-Core: Discovering and Exploring Patterns in Subspaces

Principal Investigator: Wei Wang, Jan Prins
Funding Agency: National Science Foundation
Agency Number: IIS-0812464

Abstract
High-throughput experiments have revolutionized many areas of scientific endeavor. In contrast to traditional experimental methods, they generate vast amounts of data, which is only approachable through computer-aided data analysis. The analysis is complicated by the high-dimensionality and large volume of the data. Interesting and important patterns are often embedded in unknown subspaces of the original high dimensional space. We propose to develop a series of new data mining methods that can effectively discover these subspaces, the embedded patterns, and the relationships between patterns. Our ultimate goal is to enable users to mine and explore of subspace patterns in large and complex data in an interactive rate. We propose to develop (1) new models capturing clusters or correlations in subspaces of high dimensional data. (2) efficient and scalable mining algorithms suited for large high dimensional data. (3) effective methods that model complex relationships between patterns. Our goal is to develop a framework that enables fast discovery and exploration of subspace patterns in large-scale high dimensional data. We expect our proposed methods to be applicable to a wide range of application domains to assist users to explore subspaces that contain interesting patterns and examine the relationship between patterns in an interactive fashion.