Predicting Earthquake Time Series Using Fuzzy Descriptor Models

earthquakes arrive without previous warning and can destroy a whole city in a few seconds, causing numerous deaths and economical losses. Nowadays, a great effort is being madeto develop techniques that forecast these unpredictable natural disasters in order to take precautionary measures. Two decadesago, singular systems and related fuzzy descriptor systems have been the subjects of interest due to their many practical applications in modelling complex phenomena. In this studyfuzzy descriptor models as a recently neurofuzzy interpretation of locally linear models, which have led to the introduction ofintuitive incremental learning algorithms e.g. GLOLIMOT, are implemented in their optimal structure to be compared withseveral other methods. An efficient technique, based on the error indices of multiple validation sets, is used to optimize the number of neurons as well as to prevent over fitting in theincremental learning algorithms. The scope of paper is to reveal the advantages of fuzzy descriptor models and to make afair comparison between the most successful neural and neurofuzzy approaches in their best structures and according to prediction accuracy, generalization, and computational complexity. By these modifications an accurate forecast seismic time series is obtained which is compared with several other methods.

Keywords:

Earthquake Time Series Prediction , Seismic event , Fuzzy Descriptor Models , Neurofuzzy Models , GLOLIMOT