Time Series Analysis On Mackerel (Scombridae) Landings in Malaysia
Keywords:Seasonal Autoregressive Integrated Moving Average (SARIMA) method, Multiplicative Holt-Winters method, Additive Holt-Winters method, Simple Exponential Smoothing method, Mean Square Error (MSE)
Mackerel fish is one type of pelagic fish that live in the surface of the ocean. Mackerel fish also have benefits in terms of protein which also has high demand in Asian and others countries and helps gaining profits in fisheries industries. This study will predict mackerel landings in Malaysia in one year advance which is 2018. The data of 132 monthly of mackerel landings from year 2007 until 2017 is used to make a prediction of mackerels landing by using four methods which are Seasonal Autoregressive Integrated Moving Average (SARIMA) method, Multiplicative Holt-Winters, Additive Holt-Winters Method and Simple Exponential Smoothing method. In this study, the aim is to compare the performance among four methods by measuring the accuracy of each method. The result shows that Additive Holt-Winters method is the best method used to forecast mackerel landings in 2018 as this method have the lowest value of Mean Absolute Percentage Error (MAPE) and Mean Square Error (MSE). In conclusion, the potential result from this study could be used in helping fish farmers in their annual planning of supplying fish in Malaysia.