MGT 3501 Lecture Notes - Lecture 13: Exponential Smoothing, Continual Improvement Process, Time Series
Document Summary
Identifies the relationship between two or more variables. Visually: there"s a relationship between temperature and sales of coca- Does not have to be time series. Sometimes a trend or relationship does not exist. Best technique to star with is simple moving average. Goal: identify the best fit trend line. How: find the line that has the least amount of error, by minimizing the distance between the actual values and trend line. Visualizing, can"t tell which is the best fit trendline, have to use math to find which one gives us the least error. Error = the distance between the data point and the line is the error. Need to sum up all of the error. Whichever line has the least amount of error = best fit line. Excel includes data analysis tools that perform trend lines and least square errors. R2 value is the indicator of best fit. So best fit line goes through every data point.