Abstract. Gaussian process models provide a probabilistic non-parametric modelling approach for black-box identification of nonlinear dynamic systems. The Gaussian processes can h...
The implementation of software systems with large numbers of heterogeneous components calls for a powerful design methodology. Although several such methodologies have been propos...
Learning Classifier Systems use evolutionary algorithms to facilitate rule- discovery, where rule fitness is traditionally payoff based and assigned under a sharing scheme. Most c...
Abstract. Achieving real-time response to complex, ambiguous, highbandwidth data is impractical with conventional programming. Only the narrow class of compressible input-output ma...
Genetic Programming uses trees to represent chromosomes. The user defines the representation space by defining the set of functions and terminals to label the nodes in the trees. ...