Research Research at the DEC Lab comprises interconnected developments in an unusual range of disciplines, such as software agents; machine understanding; how children learn; human and machine vision; audition; speech interfaces; wearable computers; affective computing; advanced interface design.     Research Groups Each DEC Laboratory faculty member leads a research group that includes a number of graduate student researchers and often involves undergraduate researchers.     Research Group Project and Description   Ayo: The Awari Player Mohammed Daoud, Nawwaf Kharma, Ali Haidar, Julius Popoola Awari is a two-player end-game played on a plank with 12 pits and 48 seeds; the goal of the game is to collect 25 seeds before the other player does.  In this paper, we illustrate the importance of problem domain representation, using our own Awari playing program: Ayo. We use a Genetic Algorithm to optimize the weights of the feature evaluation function of Ayo. We play Ayo against a commercially available Awari player, then compare Ayo's results to the results achieved by an older Awari player; one  that uses a 7-levels deep minimax search. Ayo, with a 5-levels deep minimax search, returns better results, due to better more intelligent representation of the state space.        Bluenome: A Novel Developmental Model of Artificial Morphogenesis Taras Kowaliw, Peter Grogono, Nawwaf Kharma The Bluenome Model of Development is introduced. The Bluenome model is a developmental model of Artificial Morphogenesis, inspired by biological development, instantiating a subset of two-dimensional Cellular Automata. The Bluenome model is cast as a general model, one which generates organizational topologies for finite sets of component types, assuming local interactions between components. Its key feature is that there exists no relation between genotypic complexity and phenotypic complexity, implying its potential application in high-dimensional evolutionary problems. The Bluenome model is first applied to a series of application-neutral experiments, in which it is shown experimentally that it is capable of producing robust agents in a reasonable amount of computation. Next, it is applied to an application involving the design of embedded agents. This second series of experiments contrasts the Bluenome model against a model in which there exists a bijective relation between genotype and phenotype, showing that the Bluenome model is capable of performing as well or better in cases of high phenotypic complexity. Additionally, genomes from the Bluenome Model are shown to be capable of re-development in differing environments, retaining many relevant phenotypic properties.     CellNet CoEv: Co-Evolving Robust Pattern Recognizers Nawwaf Kharma, Taras Kowaliw, Christopher Jensen, Hussein Moghnieh, Jie Yao An evolutionary model of classifier synthesis is presented. The CellNet system for generating binary pattern classifiers is used as a base for experimentation. CellNet is extended to include a competitive co-evolutionary mechanism, where patterns (prey) evolve as well as pattern classifiers (hunters). This is facilitated by the addition of a set of topologically-invariant camouflage functions, through which patterns may disguise themselves. The addition of the co-evolutionary mechanism allows for a) the creation of a much larger and more varied pattern database (from the original), and also b) artificially increases the difficulty of the classification problem. Application to the CEDAR database of hand-written characters yields a) an increase in the accuracy and robustness of recognition, as well as b) the elimination of over-fitting, relative to the original CellNet software.     Page 1, Page 2, Page 3, Page 4 DEC Lab Home Last Updated April 6, 2004