Most of the current Intrusion Detection Systems (IDS) examine all data features to detect intrusion or misuse patterns. Some of the features may be redundant or contribute little (...
Srilatha Chebrolu, Ajith Abraham, Johnson P. Thoma...
Stimulus selectivity of sensory systems is often characterized by analyzing responseconditioned stimulus ensembles. However, in many cases these response-triggered stimulus sets h...
We propose a feature selection method that constructs each new feature by analysis of tight error clusters. This is a greedy, time-efficient forward selection algorithm that itera...
Local pattern discovery, pattern set formation and global modeling may be viewed as three consecutive steps in a global modeling process. As each of these three steps have gained a...
Abstract. We propose a method for global validation of gene clusterings. The method selects a set of informative and non-redundant GO terms through an exploration of the Gene Ontol...