ADM 3301 Lecture Notes - Lecture 8: Seasonality, Simple Linear Regression, Standard Deviation
1 Aug 2018
School
Department
Course
Professor
Document Summary
> notice that we have two ratio-to-moving-averages for each season. To isolate the seasonal components (st), we need to identify a seasonal factor for each quarter. > seasonal factors (sfq) are obtained by averaging the ratio-to- moving-averages associated with each season in the time series. > by averaging the periods, we cancel out random component, rt , that leaves us only with seasonal component st: normalize seasonal factors (nsf) > normalized seasonal factors (nsfq) are seasonal factors that are adjusted so that they add up to the number of seasons in the time series. > what do these normalized seasonal factors tell us ? for example we can say that quarter 2 has about 14% more sales than an average quarter during the year. > remove the seasonality from the data using the seasonal components just computed. > obtain the simple linear regression results using the deseasonalized sales as the dependent variable and time as the independent variable:
