PuSH - Publication Server of Helmholtz Zentrum München

Ehler, M. ; Filbir, F.D. ; Mhaskar, H.N.*

Learning biomedical data locally using diffusion geometry techniques.

, 125-131 (2012)
DOI Order publishers version
In recent years, diffusion geometry techniques have been used in many applications of pattern classification and data visualization of high dimensional unstructured data. Our recent research has laid the mathematical foundations for query modeling on such data. We describe some features of this research, and demonstrate how the ideas can be applied fruitfully to the analysis of Cleveland Heart Disease Database and retinal multi-spectral imaging data. Even though our techniques are general purpose rather than tailored to any one application, our methods outperform such classical techniques as support vector machines in some examples.
Additional Metrics?
Edit extra informations Login
Publication type Article: Journal article
Document type Scientific Article
Keywords Data Visualization ; Diffusion Geometry ; Query Modeling
ISBN 9780889869202
Quellenangaben Volume: , Issue: , Pages: 125-131 Article Number: , Supplement: ,