International Journal of Scientific and Research Publications, Volume 3, Issue 10, October 2013 1 ISSN 2250-3153 www.ijsrp.org Demand Forecasting For Economic Order Quantity in Inventory Management Aju Mathew*,Prof.E.M.Somasekaran Nair**,Asst Prof. Jenson Joseph E*** * Mechanical Engineering Department, SCMS School of Engineering and Technology,Karukutty, Kerala ** Mechanical Engineering Department, SCMS School of Engineering and Technology,,Karukutty, Kerala *** Mechanical Engineering Department, SCMS School of Engineering and Technology, ,Karukutty, Kerala Abstract-With today’s uncertain economy,companies are searching for alternative methods to keep ahead of their competitors.Forecasts of future demand will determine the quantities that should be purchased,produced and shipped.In this work α,two data mining methods,artificial neural network(ANN) and exponential smoothing(ES) were utilized to predict the demand of the fertilizer(Ammonium Sulphate).The training data used was the sales data of fertilizer of the previous 3 years.Demand forecasted by artificial neural network is more accurate and have less inventory costs than exponential smoothing method. Index Terms- Artificial neural network(ANN),Economic Order Quantity(EOQ), Exponential smoothing(ES),Inventory Costs I. INTRODUCTION orecasting in general is prediction of some future event.Businesses use a variety of forecasts such as forecasts of technology,economy and sales of product or service.As a result,theaccuracy of demand forecasts will significantly improve the production schedulingcapacity planning,material requirement planning and inventory management.Although having accurate forecasts have never been easy,it has become more difficult in recent years due to increased uncertainty,complexity of business and reduced product life cycle.Traditionally,statistical methods such as time series analysis like exponential smoothing,weighted average,weighted moving averages,holt’s model,winter’s model etc are used for quantitative forecasting.General problems with the time series approach include the inaccuracy of prediction and numerical instability.Most of the traditional timeseriesmethods are model based which are more difficult to develop.Recently,applications of artificial neural networks have been increasing in business.One of the important applications of ANN is in the area of sales forecasting.Several distinguishing features of artificial neural networks make them valuable and attractive for forecasting tasks,artificial neural networks are data driven self adaptive method.There are a few a priori assumptions about the models for problem under study.After learning the data presented to them(a sample)ANNs can correctly infer the unseen part of the population. II. METHODOLGY Two methods used in this study were Artificial neural network method and Exponential smoothing method: A. Exponential smoothing method It calculates the smoothed series as a damping coefficient times the actual series plus 1 minus the damping coefficient times the lagged value of the smoothed series. The extrapolated smoothed series is a constant, equal to the last value of the smoothed series during the period when actual data on the underlying series are available. While the simple Moving Average method is a special case of the ES, the ES is more parsimonious in its data usage. F t+1 = α D t + (1 - α) F t where:  D t is the actual value  F t is the forecasted value  α is the weighting factor, which ranges from 0 to 1  t is the current time period. Notice that the smoothed value becomes the forecast for period t + 1. B. ANN METHOD F