Predictive coding based on non-singleton fuzzy regular networkk
基于非单点模糊正则网络的图像预测编码

A predictive coding model based on non-singleton fuzzy regular network (NSFRN) was presented. In this model, non-singleton fuzzier was introduced into regular neural network, so that noise could be removed from train data adaptively. After being simulated and compared with other predictive coding models, the model shows much better performance on anti-jamming capability, prediction precision and the quality of reconstruction pictures.