answers to your forecasting questions q. A couple of weeks ago you published the Benchmarking Report on Forecast Errors based on the data of 2004-2008. It is one of the best benchmarks on forecast accuracy done by any firm. I have one question for you, however. In the report there is no mention of the survey base: total number of customers surveyed, what countries, types of industry, etc. For example, in Food and Beverages, how many were in food service, private label, etc. Ron Burnette—Logility a. The results of this survey were based on the forecasters who attended the forecasting and tutorials events of IBF during 2004-2008. The total number of people who participated in the survey is close to 4,500. The number of people who responded to the questions on forecasting errors varied, depending on industry, type of forecast error, and forecast horizon. In total, about 800 people responded to the question on forecast errors. In the Food and Beverages, about 130 people responded. In the survey we ask them to identify the industry, which in this case is Food and Beverages, so, we don’t have any further breakdown. q. What do the residuals signify in regression? What is the reason ...

From Issue: Fall 2009
(Fall 2009)

image
Answers To Your Forecasting Questions