Missing data handling is an important preparation step for most data discrimination or mining tasks. Inappropriate treatment of missing data may cause large errors or false result...
Abstract--This paper presents a new wavelet-based image denoising method, which extends a recently emerged "geometrical" Bayesian framework. The new method combines three...
Aleksandra Pizurica, Wilfried Philips, Ignace Lema...
In this paper, we investigate an approach based on support vector machines (SVMs) for detection of microcalcification (MC) clusters in digital mammograms, and propose a successive ...
Issam El-Naqa, Yongyi Yang, Miles N. Wernick, Niko...
In this paper, we present the results of our work that seek to negotiate the gap between low-level features and high-level concepts in the domain of web document retrieval. This wo...
We present a method for sampling feature vectors in large (e.g., 2000 5000 16 bit) images that finds subsets of pixel locations which represent "regions" in the image. Sa...