Analytics with Simple Regression to Identify Drivers and Forecast
摘要
Analytics from regression can easily create a long range forecast based on trend. Regression and correlation, upon which regression is based, reflect linear association between two variables. Regression quantifies the influence of a continuous, independent driver x on a continuous dependent, performance variable y. In the case of a trend focused forecast, the driving variable x is time period. In later chapters, focus will be expanded to both explain how an independent decision variable x drives a dependent performance variable y and also in predicting performance y to compare the impact of alternate decision variable x values. X is also called a predictor, since from x we can predict y. Here, focus is on prediction of performance or response y from knowledge of the driver, time period, x.