This article proposes the use of two evolutionary algorithms (EAs) to the dynamic difficulty adjustment (DDA) of a serious game in the rehabilitation robotics application. DDA occurs in runtime for a better user experience with a game. This approach is used to improve the quality of the game experience and to avoid boredom or frustration for players with severe limitations imposed by pathologies such as stroke, cerebral palsy, and spinal cord injuries. The first EA solves the game adjustment problem, changing the game difficulty according to the player’s skill, and the purpose of the second EA is to adjust the coefficients of the first EA’s objective function so that it can work in a more effective way. To do so, the second EA uses results of game matches against simulated player profiles. The results shows that the presented method was able to identify a set of coefficients that allows the first EA to correctly adjust the difficulty level for all six tested player profiles.
AgradecimentosDeixo aqui meus agradecimentos a algumas pessoas valiosas que contribuíram de maneira direta ou indireta na conclusão deste trabalho. Ao professor Mário de Castro Andrade Filho que muito me auxiliou com o método estatístico utilizado neste trabalho, muito obrigado pela sua paciência e boa vontade em me ajudar. Ao meu orientador e amigo Eduardo do Valle Simões, que me guiou neste projetos e também em anteriores que muito contribuíram com este trabalho. Ao meu grande amigo Francisco Krug, por compor e produzir a excelente música utilizada no jogo criado neste trabalho. À querida Luziana Sant'Ana Simões pela ajuda com a revisão ortográfica e por acompanhar todos os meus trabalhos de perto há muito tempo, me motivando a continuar sempre. A todos de minha família, em especial a Francisco Constantino Crocomo, Faridi Kassouf Crocomo, Lucas Kassouf Crocomo, Khalil Kassouf e Selma Khalil Kassouf, por estarem sempre me apoiando e torcendo por mim. Para todos àqueles que leram este trabalho e deixaram suas valiosas contribuições, e também aos que sempre perguntaram sobre o andamento deste projeto: obrigado, o interesse demonstrado por vocês foi responsável pela qualidade que pude atingir neste trabalho. Espero que gostem do resultado final! O presente trabalho foi realizado com o apoio do CNPq, uma entidade do Governo Brasileiro voltada ao desenvolvimento científico e tecnológico. PUBLICAÇÕES REFERENTES A ESTE TRABALHO CROCOMO, M. K.; MIAZAKI, M.; SIMÕES, E.
Estimation of Distribution Algorithms (EDAs) have proved themselves as an efficient alternative to Genetic Algorithms when solving nearly decomposable optimization problems. In general, EDAs substitute genetic operators by probabilistic sampling, enabling a better use of the information provided by the population and, consequently, a more efficient search. In this paper the authors exploit EDAs' probabilistic models from a different point-of-view, the authors argue that by looking for substructures in the probabilistic models it is possible to decompose a black-box optimization problem and solve it in a more straightforward way. Relying on the Building-Block hypothesis and the nearly-decomposability concept, their decompositional approach is implemented by a two-step method: 1) the current population is modeled by a Bayesian network, which is further decomposed into substructures (communities) using a version of the Fast Newman Algorithm. 2) Since the identified communities can be seen as sub-problems, they are solved separately and used to compose a solution for the original problem. The experiments showed strengths and limitations for the proposed method, but for some of the tested scenarios the authors’ method outperformed the Bayesian Optimization Algorithm by requiring up to 78% fewer fitness evaluations and being 30 times faster.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.