Self-organizing mixture networks for probability density estimation
IEEE Transactions on Neural Networks
Hi-index | 0.01 |
A multiscale self-organizing mixture network (MSOMN) is proposed for learning mixture multiscale autoregressive model of synthetic aperture radar (SAR) imagery. The MSOMN combines the multiscale method, the Kullback-Leibler information metric, the stochastic approximation method, and the selforganizing map structure. Updating of the parameters is limited to a small neighborhood around the winner that is based on maximum posterior probability. The network possesses a simple structure, and yields fast convergence, which is confirmed by experimental results of SAR images.