Thesis (BS Applied Mathematics) -- University of the Philippines Mindanao, 2011

This study introduced the use of Pareto principle in considering the number f perishable products offered by one of the biggest company in Davao City that should be included. It has searched the capability of ARIA models to produce good forecasts of the product sales incorporating their shelf lives. The analysis was only focused on the number of quantity of the products due to the confidentiality that the company is holding. The existing daily time series data form January 01, 2009 to September 30, 2009 was used as the testing sample to develop models that give the best fitting model, and the period from October 01, 2009 t December 31, 2009 was used as the holdout sample for each product. Forecasts were then modified by noting shelf lives of the products to test the validity and efficiency of the generated model. Number of wastages was also considered in comparing the existing method of the company and the proposed method generated. Results showed that 119 out of 126 models generated by the proponent found to be better than the existing method the company is using. Based on the overall results, the proposed method is better than the existing one, and the proposed models will be a big help to the company being studied in minimizing the number of returns.