FORECASTING AND SIMULATION

Forecasting and Simulation

Forecasting and Simulation

Description of Data

Data that has been examined in this research represents the manufacturers' new orders for electronic and other electrical equipments. Quarter-wise data have been collected from year 1983 to year 2007.

The data represent scale item variable for the number of manufacturing orders that placed in every quarter for the defined period from year 1983 to 2007. Company production is highly dependent on the number of manufacturing orders it receives from its supply chain partners. Manufacturers' new orders data only categorize the orders for the product having model number N2422 that are required in the production of electronic and electrical equipments. Time Series Plot of manufacturing orders for equipments shows continuous variation in the level. Numbers of manufacturer's orders have changed continuously in each quarter. This variation in the data shows the component of uncertainty that the company should address in order to effectively manage the supply of product N2422 on placement of new orders by manufacturers. Historical data for output show an inclining trend over the period from Year 1983 to Year 2007. Diagram given below shows the time series plot of the manufacturers' order for electronic and electric equipments.

Forecasting Techniques

Future demand for the product is extremely variable to the company, which is beyond their control. The appropriate demand forecasting is important because it will enable the company to take appropriate actions based on the forecasting manufacturer's orders. Forecast is essential for any knowledge of the historical demand of the product. The demand forecasting future decisions may be taken in devising the policy of purchases of raw materials and other items, the size of the production lots to build, stock levels in the warehouse and safety stock, and frequency of orders for the product (David, 2006). Therefore, in order to assess the trend of the given output, different approaches can be applied depending upon the future prospects. The data selected involve time factor which includes seasonal and dimensional changes in productivity level.

The purpose of the non-causal forecasting techniques is to obtain estimates or forecasts of future values of a time series from the historical information contained in the observed series to date (Creswell, 2009). These techniques do not require specification of the factors that determine the behavior of the variable, but based solely on modeling the systematic behavior of the series. The appropriate technique depends on the model prediction of behavior of the series. The assumptions underlying the forecasting method selection is the stability to assess the systematic behavior of the series (Silverman, 2004). Secondly, the value of the variable observed in any period is the result of systematic behavior and a random disturbance (Kevin, 2003).

Selected Models

Three techniques that are suitable to assess the forecasting pattern of the selected data include Expert Modeler, Exponential Smoothing, and Seasonal Moving Average Model.

Discussion and Justifications for the Selected Models

Expert Modeler

The data selected involve time factor which includes seasonal and dimensional changes in orders level. Considering the impact of seasonality and trend component for the given ...