The strong convergence of visual classification method and its applications

Visual classification method has been proposed as a learning strategy for pattern classification problem. In this paper, we show the strong convergence property of this method. In particular, the method is shown to converge to the Bayesian estimator, i.e., the learning error of the method is convergent to the posterior expected minimal value. The performance of the method has also been theoretically evaluated to comply with the human visual sensation and perception principle. The method is successfully used to some practical remote sensing and disease diagnosis applications. The experimental results all verify the validity and effectiveness of the theoretical conclusions.