Support Vector Machines (SVMs) have been successfully applied to solve a large number of classification and regression problems. However, SVMs suffer from the catastrophic forgetti...
Abstract. Relationships are an essential part of the design of a database because they capture associations between things. Comparing and integrating relationships from heterogeneo...
Abstract. This paper investigates the methods for learning predictive classifiers based on Bayesian belief networks (BN) – primarily unrestricted Bayesian networks and Bayesian m...
This paper uses the GP paradigm to evolve linear genotypes (individuals) that consist of Java byte code. Our prototype GP system is implemented in Java using a standard Java devel...
We study the definability of constraint satisfaction problems (CSP) in various fixed-point and infinitary logics. We show that testing the solvability of systems of equations over...