Prediction of Fishing Ground based on RBF Neural Network

This study tested Radial Basis Function Neural Network (RBFN) as an intelligent method to fulfill the prediction of fishery forecasting in Southwest Atlantic on Illex argentines. Due to the existing drawback of fuzzy C-means (FCM) RBF which is time consuming, we used symmetry-based Fuzzy C-means (SFCM) to improve the effectiveness of RBF. Altogether Six marine environmental factors are considered which are months, longitude and latitude, sea surface temperature (SST), Sea surface Height (SSH) and chlorophyll for predicting the Habitat Suitability Index (HSI). The traditional calculation methods of HSI are statistical ways such as multiple linear regressions. The results obtained from the SFCM/RBF model were compared with Multiple Linear Regressions in terms of accuracy criterions MSE, RMSE. Through the prototype system, it is shown that the intelligent model has high predictive ability and better goodness of fit compared with statistical models.