Distributed dynamic data driven prediction based on reinforcement learning approach

Abstract

In this paper, we propose a new distributed dynamic data driven model and strategy to direct and evaluate the interlinked data sets in Dynamic Data Driven Application Systems (DDDAS). The underlying technique is the introduction of a reinforcement Q-Learning approach including search strategies to determine how to drill and drive a series of highly dependent data in order to increase prediction accuracy and efficiency. In simulation, the new model utilizes individual sensors, distributed databases, and predictors in Dynamic Data Stream Nodes with multiple dimensional variables which can be instantiated to explore the search space, so that search convergence can be improved. We show the effectiveness and applicability of using the technique in the analysis of typhoon rainfall data. The result shows that the proposed approach performed better than traditional linear regression approaches, reducing the error rate by 30.48%.