In the feature selection of cancer classification problems, many existing methods consider genes individually by choosing the top genes which have the most significant signal-to...
In this paper, we present a novel information theoretic approach to image segmentation. We cast the segmentation problem as the maximization of the mutual information between the ...
Alan S. Willsky, Anthony J. Yezzi Jr., John W. Fis...
—In this paper we present an Information Theoretic Estimator for the number of sources mutually disjoint in a linear mixing model. The approach follows the Minimum Description Le...
Random data perturbation (RDP) has been in use for several years in statistical databases and public surveys as a means of providing privacy to individuals while collecting informa...
The design of software is often depicted by graphs that show components and their relationships. For example, a structure chart shows the calling relationships among components. O...