How to Choose among Three Forecasting Methods: Machine Learning, Statistical Models, and Judgmental Forecasts
In this article, Yue Li, Diane Berry, and Jason Lee of the Global Advanced Analytics Group with Bain & Company present their views on the strengths and weaknesses of the three forecasting methods.
“Forecasting methods are usually categorized into three types: statistical models, machine-learning models, and judgmental (or expert) forecasts. Based on their forecasting experiences across many industries, the authors offer recommendations on how to decide when to use each model as well as when a combination of methods should be considered. In the Commentary that follows, Stephan Kolassa presents his own assessment of the interpretability, stability, and scalability of the three different approaches, in the process clarifying the bases for these ratings.”
Download the full article, with our compliments!
Foresight is a quarterly publication of the International Institute of Forecasters. The premier journal for working forecasters and planners, Foresight delivers peer-reviewed and edited articles, tutorials, and case studies that provide unmatched professional development for forecasters. Become a member now and receive Foresight!