A Comparison Of Demand Forecasting Models

1877 Words8 Pages
Ibrahim Aljassim Data Set A “A Comparison of Demand Forecasting Models” INTRODUCTION Blattberg and Hoch have stated that (forecasting) “remains an art with tenuous scientific superstructure.” Despite this claim, numerous time series models have been created that can provide significantly more accurate forecasts for future demand than simply ‘going with your best guess.’ If something is not measured, it will never improve, or stated otherwise, you ‘get what you inspect, not what you expect.’ Time series models seek to have an accurate and unbiased forecast. This forecast than can be used to both avoid the waste, in inventory and storage costs, of over-forecasting and the cost of lost sales and unhappy customers due to under-forecasting. The basics of each type of time series model are very similar. Each type uses actual data to factor out the randomness of the data and to determine if there are trends and/or seasonality in the data. The advantages of time series models are much more predictable budgeting for materials, labor, sales, costs, and other factors that drive the profit margin of a successful business. Also, they are very easy to use and cheap to implement. The disadvantages come into play when the forecasting models are not accurately applied or are not trusted sufficiently to gain the advantages that these models can provide, and one of the main disadvantages is that these type of models just look back and give you predictions so it all based
Open Document