A Hybrid Rough Fuzzy Approach Aim to Solar and Geomagnetic Storms Forecasting

This paper illustrates using hybrid rough fuzzy approach for developing a data-driven model of the alert system. In this research, a reliable alert system for prediction of different situations and setting off of the alerts for different human industry particles is designed. Preprocessing and reduction of data with data mining methods is performed, then rough sets theory are used as an important data mining method to attain the regular and reduced information from the previous activities of the system. Finally, using the produced and reduced rules obtained from rough sets reduction algorithms, the fuzzy inference system is devised to predict and setting off of suitable alerts. By using of this approach for handling and modeling of real world phenomenal, our experiments is indicated that the proposed approach can use for handling uncertainties and reach a suitable performance in compare with other prediction methods when we have several certain features set for representing the knowledge.

Keywords:

Data Mining , Rough Sets theory , Fuzzy Systems , Data-Driven Modeling , Space Weather , Alert Systems