TIME SERIES ANALYSIS  OF EXPORT FLAT BED SHEET AND PILLOW  PRODUCT:  INCASE OF KOMBOLCHA TEXTILE SHARE COMPANY

ALEMAYEHU, AMSALU

dc.contributor.author	ALEMAYEHU, AMSALU
dc.date.accessioned	2017-10-23T10:20:17Z
dc.date.available	2017-10-23T10:20:17Z
dc.date.issued	2017-10-23
dc.identifier.uri	http://hdl.handle.net/123456789/8097
dc.description.abstract	Textiles along with clothing sector are truly considered t he backbone of the world’s economy. The main objective of this study was Time Series Analysis of the export flat bed sheet and pillow product of Kombolcha Textile Share Company . The descriptive and inferential statistical techniques were used for data analysis in this study. Univariate Time series analysis uses only the past history of the time series forecast plus current and past random error terms. Time series data used in this analysis consists of the monthly data of 2007 to 2016 on export flat bed sheet and pillow product univariate secondary data obtained from the monthly report of the Kombolcha Textile Share Company. The Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) were the most important elements for model identificatio n and AIC and SBC were also used to select the best fit of model adequacy. Forecasting is the prediction of a future value Y n +k from a series of n previous values Y 1 , Y 2 , . . . , Y n . The most common measures of predictive accuracy were MSE, RMSE, and MAE. The ACF and PACF for the export flat bed sheet and pillow product were large negative significant spike at lag 1 (this means that the ACF and PACF of the successive pairs of observations within 1 time period is not within sampling error of zero). All of the other ACF and PACF for lags 2 to 30 are within the 95% confidence limits. ARIMA(1,1,1)with corresponding AR(1) and MA(1) were best fit of the Box Jenkins model or in the mean model ARCH(1)and GARCH(1,1) were also best in the variance model in this study. The modified Box-Pierce (Ljung-Box) chi-square statistics of 6.5, 12.7, 19.2, and 20.6 give p-values of 0.688, 0.920, 0.974, and 0.999 show that there is good fit of the model. Because the p-values are quite large (greater than 0.05), since we conclude that the residuals appear to be uncorrelated. The minimum value of forecasting performance, MAPE=40.3 is better fit of the model forecasting accuracy. The scaling of Thiel’s inequality coefficient U=0.165 was close to zero, it is good predictive performance of the model.	en_US
dc.language.iso	en_US	en_US
dc.subject	Statistics	en_US
dc.title	TIME SERIES ANALYSIS OF EXPORT FLAT BED SHEET AND PILLOW PRODUCT: INCASE OF KOMBOLCHA TEXTILE SHARE COMPANY	en_US
dc.type	Thesis	en_US