| Resumen |
The aim of evolutionary vision is to address typical problems of artificial vision through techniques whose principles are based on the theory of biological evolution. This allows us to consider the visual problem as a goal-oriented vision problem. Although the current automatic approaches are widely used in diverse recognition tasks, these are unable to explain how the knowledge to solve the problem is derived. In this paper, we use an evolutionary vision technique called brain programming (BP) to extract the knowledge used for the classification task. This model allows us to know how the knowledge to solve the leukemia images classification problem is derived. In addition, we present two variants focused on transferring knowledge to improve the classification. Results show that classification on leukemia images is achieved successfully from the solutions obtained by the proposed variants. |
