Considerations for the Development of Evidence Theory Applications

Abstract

Evidence theory provides a natural framework for analyzing imprecise probabilities. Evidence theory enables the explicit recognition of ignorance and is therefore suited for the analysis of missing or incomplete data. This paper examines the characteristics of uncertainty beyond traditional probability theory, and is motivated by three primary objectives: (i) introducing appropriate roles for uncertainty theories and their associated relevance, (ii) developing guidance and methods for data combination applications using evidence theory, and (iii) developing evidence theory applications within structural design. The paper reviews the development and applications of evidence theory in order to motivate further development for specific practical applications. Potential structural design applications for evidence theory include dynamic loading and performance-based design. The continued development of data combination methods using evidence theory can reduce the additional efforts required to apply evidence theory and make future evidence theory applications more efficient and effective.