Jointly Modeling Label and Feature Heterogeneity in Medical Informatics
Abstract
Multiple types of heterogeneity including label heterogeneity and feature heterogeneity often co-exist in many real-world data mining applications, such as diabetes treatment classification, gene functionality prediction, and brain image analysis. To effectively leverage such heterogeneity, in this article, we propose a novel graph-based model for Learning with both Label and Feature heterogeneity, namelyL2F. It models the label correlation by requiring that any two label-specific classifiers behave similarly on the same views if the associated labels are similar, and imposes the view consistency by requiring that view-based classifiers generate similar predictions on the same examples. The objective function forL2Fis jointly convex. To solve the optimization problem, we propose an iterative algorithm, which is guaranteed to converge to the global optimum. One appealing feature ofL2Fis that it is capable of handling data with missing views and labels. Furthermore, we analyze its generalization performance based on Rademacher complexity, which sheds light on the benefits of jointly modeling the label and feature heterogeneity. Experimental results on various biomedical datasets show the effectiveness of the proposed approach.
Document Details
- Document Type
- Pub Defense Publication
- Publication Date
- May 11, 2016
- Source ID
- 10.1145/2768831
Entities
People
- Dawei Zhou
- Haoda Fu
- Hongxia Yang
- Jieping Ye
- Jingrui He
- Pei Yang
- Theodoros Lappas
Organizations
- Arizona State University
- Defense Advanced Research Projects Agency
- Eli Lilly and Company
- Jerry and David's guide to the World Wide Web
- National Natural Science Foundation of China
- National Science Foundation
- Stevens Institute of Technology
- United States Army Research Laboratory
- University of Michigan