查看文章

Current neural network models, mostly concerned with bottom-up processes, such as finding optimal parameters for a given set of data (which correspond to the stimuli of experiments), do not incorporate top-down information, such as preselecting features or internal knowledge. New experimental results, however, show that attention and other higher cortical processes play an important role in perceptual learning issues. After briefly reviewing current (mathematical) learning models, we present these new results and sketch out a framework of perceptual learning that takes top-down influences into account.