Forecasting of monthly marine fish landings using artificial neural network

Abstract

Management of marine resources have gradually become more important during these past years because of the increased awareness of these resources becoming limited. Forecasting of fish landings is one of the many ways that can contribute to a better decision making for fisheries management. Being a renowned forecasting model, artificial neural network with back propagation was selected for this research with enhancement made by pre-processing the data using empirical mode decomposition. The monthly marine landings data of East Johor and Pahang which has 144 observations each, was gathered from the Department of Fisheries Malaysia website. A ratio of 92:8 was used to divide the data into training and testing sets. Data pre-processing was done in R software whereas the forecasting models were developed in MATLAB software. Results from the proposed model are then compared to a conventional artificial neural network using the root-mean-square error and mean absolute error values, wherein it was shown that the proposed model could outperform the conventional model.