Predicting marine capture fish volume and its selling price, case of coastal area in Java Island, Indonesia

• Main Authors: Suwondo, Endy, Yuliando, Henry, Muchfirodin, Muchamad
• Format: Conference Proceeding
• Language: English
• Subjects: Artificial neural networks; Coastal environments; Coasts; Correlation analysis; Data processing; Fish; Fisheries; Neural networks; Regression analysis; Regression models; Vessels
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/1.4958559

This research aims to develop a model for predicting marine capture fisheries and its selling price as the main activity at fish auction (TPI) around Java Island, Indonesia. Data processing was started by analyzing ANOVA test to determine the characteristics differences of capture fish between the north and south of Java coastal area. Furthermore, Spearman correlation analysis was performed to determine linkage between influencing factor to yield. A regression model was constructed and combined with artificial neural network (ANN) to predict which factor that is significantly influencing the capture fish volume and the determination of selling price under auction mechanism. Based on the result, captured fish volume both of north and south coastal area of Java are dominantly affected by a number of vessels. On the other side, there is difference in factors that affect the selling price. In Java north coastal zone, the selling price is determined by the amount of capture fish and number of vessels, while in south, the selling price is strongly determined by “raman.”