FORESIGHT, Issue 53
$0.00 – $45.00
- Will You Become a Victim of Your Models? by Thomas Willemain
In this feature section, Tom Willemain provokes a good deal of thought about the role of statistical models in supply chain forecasting, a field that he believes lags far behind finance in embracing algorithms over gut instinct. His article is followed by Commentaries from practitioners and researchers about the realities that can lead to ”model failure” and the conditions for successful implementation of model-based decision making.
Predicting Medical Risks and Appreciating Uncertainty by Spyros Makridakis, Ann Wakefield, and Richard Kirkham
In this paper, the authors present a number of specific examples related to risk and uncertainty in the context of clinical decision making—some more than a little alarming—including extremely high incidences of misdiagnosis, reluctance on the part of medical professionals to abandon treatment regimens that are doing patients no good and may be causing harm, and systemic flaws in medical research methodology that can impede important new data from reaching practitioners.
Includes commentaries by John P.A. Ioannidis and Nassim Nicholas Taleb
The Ten Commandments of Economic Forecasting by Azhar Iqbal and John Silvia
In this ever-evolving world, we need accurate forecasts of key economic and financial variables to help decision makers design effective policies. Azhar Iqbal and John Silvia present a framework of “ten commandments” to govern the economic-forecasting process. They begin with the need to know the forecasting objective and loss function and end with the monitoring of forecast performance and refinements to the original model.
- Monetized Forecast-Error Comparisons by Shaun Snapp
Standard error metrics such as the MAPE are not effective in directing a company to make changes that will have the most significant impact. Rather, it is in the monetization of forecast error that we enable the best use of forecasting resources. This article illustrates the calculation of the monetized forecast-error metric and presents the author’s free application designed to steer improvement in forecast accuracy at specific product/locations.
Complete Issue, Special Feature 1, Special Feature 2, Special Feature 3, Article 1, Article 2, Article 3, Article 4, Free Article 1, Free Article 2, Free Article 3, Free Article 4