Simple versus Complex Forecasting Models (April 14)

1 post / 0 new
Simple versus Complex Forecasting Models (April 14)


The following information is being sent to the TMIP listserv, with the permission of the hosts for this “live stream” session. In other words:  I have no involvement with the session!

While the presentation is about “population projections” rather than “travel forecasting,” based on what’s in the abstract, it could have some interesting content:


George Washington University’s H. O. Stekler Research Program on Forecasting (, joint with the Federal Forecasters Consortium (FFC) (, is pleased to invite you to our next forecasting seminar:

“Simple versus complex forecasting models: thinking through their pros and cons."

Live streaming: Thursday April 14 2022, 12:30 pm - 2 pm ET:
Anne Morse (Census) will present.

Abstract: There is an intuition that complex, computationally sophisticated projection methods produce more accurate projections than simple methods. This intuition is so strongly held, in fact, that it persists despite evidence to the contrary. This talk will discuss under what conditions we might expect simple projection models to outperform complex ones and vice-versa.  The talk gives a particular emphasis to population projections but the themes are broadly applicable to all forecasting fields.

Live stream: Join Zoom meeting using the following link:
• passcode: 042414