The testing and formal verification of black box software components is a challenging domain. The problem is even harder when specifications of these components are not available...
In political speeches, the audience tends to react or resonate to signals of persuasive communication, including an expected theme, a name or an expression. Automatically predicti...
This paper proposes a novel method for blindly separating unobservable independent component (IC) signals based on the use of a genetic algorithm. It is intended for its applicati...
Unsupervised learning algorithms aim to discover the structure hidden in the data, and to learn representations that are more suitable as input to a supervised machine than the ra...
Abstract Clustering Stability methods are a family of widely used model selection techniques for data clustering. Their unifying theme is that an appropriate model should result in...