Forecasting Modelling of Cockles in Malaysia by Using Time Series Analysis


  • Hani Nabihah Aziz
  • Mohd Saifullah Rusiman
  • Siti Noor Asyikin Mohd Razali
  • Abdul Wahab Abdulla
  • Nur Amira Azmi





Auto regressive moving average (ARMA), Holt-Linear, Mean Square Error (MSE


Cockle farmed in Malaysia are from Anadara genes and Arcidae family which known as blood cockle. Normally, it was found in the farmed around mangrove estuary areas in the muddy and sandy shores. This study aims to predict the production of cockle to ensure sure the cockle supplies are synchronised with the demand. Then, based on the demand, the prediction result could be used to make decision either to import or export the cockle. The data were taken from the Department of Fisheries Malaysia (DFM) and it has cyclic pattern data. There are two methods used in this study which are Holt-Linear method and Auto regressive moving average (ARMA). In determining the best fitted model between the two methods, the mean square error (MSE) values will be compared and the lowest value of MSE will assign as the best model. Result shows that ARMA(1,1) is the best model compared to Holt-Linear. Therefore, ARMA(1,1) model will be used to forecast the production of cockle in Malaysia.


[1] Tookwinas S, Commercial Cockle Farming in Southern Thailand, University of California, International Center for Living Aquatic Resources Management (1985).

[2] Poutiers JM, Bivalves (Acephala, Lamellibranchia: Pellecypoda) In: The living Marine Resources of the Western Central Pacific, Volume 1, Seaweeds, corals, bivalves and gastropods (Carpentier, K. E., Niem, V. H. eds), FAO, Rome, Italy, (1998).

[3] Ng FO, “Cockle Culture SAFIS Extension Manualâ€, Southeast Asian Fisheries Development Center, Bangkok 13 (1998), pp. 22.

[4] Broom MJ, The Biology and Culture of Marine Bivalve Molluscs of the Genus Anadara, ICLARM Studies and Reviews 12, International Center for Living Aquatic Resources Management, Manila, (1985), pp. 37.

[5] Bako HY, Rusiman MS, Kane IL & Matias-Peralta HM (2013), Predictive modeling of pelagic fish catch in Malaysia using seasonal ARIMA models, Agriculture Forestry and Fisheries, Vol.2, No.3, pp. 136-140.

[6] Lee MH, Rahman NHA, Suhartono, Latif MT, Nor ME, Kamisan NAB, Seasonal ARIMA for forecasting air pollution index: A case study, American Journal of Applied Sciences, Vol.9, No.4 (2012), pp. 570-578.

[7] Nor ME, Safuan HM, Shab NFM, Asrul M, Abdullah A, Mohamad NAI, Lee, MH, Neural network versus classical time series forecasting models, AIP Conference Proceedings, Vol.1842, (2017), pp. 030027.

[8] Khalid K, Mohamed I & Abdullah NA, An Additive Outlier Detection Procedure in Random Coefficient Autoregressive Models AIP Conference Proceedings 1682 050017, (2015).

[9] Mohamed I, Khalid K & Yahya MS (2016), Combined Estimating Function for Random Coefficient Models with Correlated Errors Communications Iin Statistics—Theory and Methods, Vol.45, No.4, (2016), pp. 967-975.

[10] Hyndman RJ & Athanasopoulos, Forecasting: Principle and practice, Otexts, (2014).

[11] Guo YH, Shi XP & Zhang XD, A study of short term forecasting of the railway freight volume in China using ARIMA and Holt-Winters models, In 2010 8th International Conference on Supply Chain Management and Information, (2010), pp.1-6.

[12] Veiga CPD, Veiga CRPD, Catapan A, Tortato U & Silva WVD, Demand Forecasting in Food Retail: A Comparison Between the Holt-Winters and ARIMA Models, WSEAS Transactions on Business and Economics,Vol.11 (2014), pp.608-614.

View Full Article: