Malaysian Torpedo Scads (Ikan Cincaru) Forecasting Using Time Series Analysis

Abstract

Forecasting, which can be done in a number of ways, such as statistical methods and machine learning techniques, involves making predictions about future values based on historical data. This study presents a thorough investigation into the forecasting of Malaysian Torpedo Scads catch data using time series analysis, addressing the crucial requirement for accurate predictions in the context of fisheries management and sustainable harvesting practices. The study primarily employs statistical methods, specifically the Box-Jenkins, exponential smoothing, and Hybrid methods using SPSS software, to model the temporal patterns inherent in the catch data. Given the seasonal periodic fluctuations observed in monthly Malaysian Torpedo Scads (Ikan Cincaru) data, IBM SPSS software is used to forecast catch data for 2022 utilizing the Seasonal Autoregressive Integrated Moving Average (SARIMA) approach, Simple Seasonal Exponential Smoothing method, and Simple Seasonal Exponential Smoothing method through applying hybrid data. A dataset comprising 132 observations from January 2011 to December 2021 is utilized to build the models. Among the three methods, Exponential Smoothing emerges as the most effective model, with the Simple Seasonal Exponential Smoothing exhibiting the lowest values for as Root Mean Square Error (RSME), Mean Absolute Percentage Error (MAPE), and Mean Absolute Error (MAE)—389.805, 11.751%, and 271.024, respectively.