The Application of ANN-Linear Perceptron in the Development of DSS for a Fishery Industry

Overfishing is a global environmental problem that risks fisheries since many of the fish stock of the fisheries have already reduced to below a tolerable level. One of solutions that often implemented in the fishery management is by calculating the value of Maximum Sustainable Yield (MSY) as the maximum tolerable harvest that can be taken out from the natural stock without harming the population over an indefinite period of time. A proper tool used for computing the MSY is needed to support the fishery manager in solving this decision making problem. In this paper we propose a software development of Decision Support System (DSS) to address such fishery industry problem. The DSS is developed to compute the MSY from the annual yield-effort data of the fishery. We use two sigmoid growth equations, Logistic and Gompertz equations, as the underlying population models, which then are approximated by their discrete forms for computing several growth parameters. Most known methods of growth parameter estimation use a Multiple Linear Regression with Ordinary Least Square method (MLR-OLS). Here we propose the application of Artificial Neural Network with Linear Perceptron method (ANN-LP). A case study in this paper shows that the effectiveness of the proposed ANN-LP is as good as the MLR-OLS in estimating both the growth parameters and the MSY of the fishery in the case study needed in the computation of its MSY by using Artificial Neural Network (ANN) Linear Perceptron.