ABSTRACT
This chapter contains papers presented in a session on “Computational Approaches to Artificial Intelligence" at the International Conference of Computational Methods in Sciences and Engineering. The papers propose new globally convergent first-order batch training algorithm for neural networks and focuses on knowledge representation methods presenting compositions of clones of Boolean functions. It explores nonlinear systems theory as applied to modeling the emergence of economic agglomerations, addresses the problem of integrating user preferences with network Quality of Service parameters for the streaming of media content, and suggests protocol stack configurations that satisfy user and technical requirements to the best available degree. The papers also introduce Unified Particle Swarm Optimization, a new scheme that harnesses the local and global variants of the standard Particle Swarm Optimization algorithm, and consider data fitting schemes that are based on different norms to determine the parameters of curve–models that model landslides in dams.
