The objective of this research is to construct the appropriate forecasting model for the export quantity of concentrated latex by the three time series analysis methods which are Box-Jenkins method Winters’ multiplicative exponential smoothing method and combined forecasting method. Time series data from the website of Office of Agricultural Economics during January 1998 to February 2014 (194 values) are used and divided into two series. The first 182 values from January 1998 until February 2013 for the models prediction and the last 12 values from March 2013 until February 2014 for comparison the performance of forecasting models via the criterion of the lowest root mean squared error and the criterion of the highest correlation coefficient between existing values and forecasted values. Research findings indicated that for all forecasting methods that have been studied the most powerful method is the combined forecasting and the forecasting model is where and represent the single forecasts at time t from Box-Jenkins method and Winters’ multiplicative exponential smoothing method respectively.

The objectives of this research were to select the most appropriate forecasting model for uncertain demand of hot-rolled steel and to find the purchase amount and the reorder point of some raw materials in order to minimize the total cost of production in an iron and steel production planning. The three forecasting models which were exponential model ARIMS model and the proposed Bayesian model were studied. An iron and steel factory in Samutprakarn province was a prototype for this study. Because of their uncertain high demand the Steel H-Beam Steel plate steel round bar and black steel round bar with various sizes were selected and their parameters including demand rate inventory cost back order cost and setup cost were collected. The study found that the proposed Bayesian forecasting model is the most appropriate. The predicted demand and all parameter values were used to find the purchase amounts and the reorder points of the raw materials under the (rQ) policy of the stochastic inventory model.