Dempster-Shafer Theory as an Applied Approach to Scenario Forecasting Based on Imprecise Probability

Scenarios are detailed descriptions of a system's possible future. Unlike the traditional method that a set of scenarios is presented without quantifying any degree of confidence or likelihood, imprecise probability is introduced to solve the scenario-forecast problem. This allows us to define upper and lower bound, a kind of degree of uncertainty, for any given future with its belief function and plausibility function derived from Dempster-Shafer Theory. Then Zhang's rule of combination is introduced as an applied approach to improve decision making based on a set of scenarios all hold equally credible but inconsistent with beliefs about the future.