Overview of data store management for sliding-window learning using MLP networks

This paper presents an overview of sliding-window based learning with data store management (DSM) techniques using multilayer perceptron (MLP) neural network. The paper views several DSM techniques used to reduce the correlation of data inside the window store. The sliding window (SW) training is a form of higher order instantaneous learning strategy without the need of covariance matrix, usually employed for modeling and tracking purposes. Sliding-window framework is implemented to combine the robustness of offline learning algorithms with the ability to track recursively the underlying process of a system. This paper view the performance of sliding window backpropagation (SWBP) with application of data store management e.g. simple distance measure, angle evaluation, weighted distance measure, weighted angle evaluation and the novel prediction error displacement. The simulation results show that the best convergence performance is gained using store management techniques.