T.U.E.S.day lecture: What to do in the absence of (good data)?
Humans frequently need to make decisions under uncertainty, albeit professionally or their private life. Sometimes the uncertainty can be accurately described by making use of existing data, but in many instances the data at hand is not representative or fraught with problems or simply not available. For example, World Health Organization (WHO) has been interested in forecasting the percent of Indian or Chinese population that will be resistant to antibiotics in 2020, whereas there is scarce or no historic data available. The Nuclear Regulatory (NRS) in U.S. has put effort in predicting possible malfunctions of chemical installations for the accident consequence management for nuclear power plants where there was little or no data available.
The answer is to perform a structured expert judgment elicitation. Expert elicitation is a powerful method to obtain quantitative estimates of uncertainty in the absence of data. I will present the Classical Model (CM), which is embedded in a mathematical framework and provides the most transparent method of performing (structured) expert judgement. The method has been developed at TU Delft more than 25 years ago. I will present some of the most important, as well as ongoing applications.
Want to enjoy a free lunch during this lecture? Register wisv.ch/lectureRegister now