DS Evidence Theory Case Studies: Implementation Examples for Data Fusion Decision Making

This article aims to explain the application of DS evidence theory in data fusion decision-making, while providing detailed case studies, implementation explanations, and references to support the theory's validity and reliability. DS evidence theory is a mathematical framework that assists decision-making by combining evidence from uncertain information sources. The theory demonstrates significant advantages in data fusion applications, particularly in integrating data collected from multiple sensors. Such data typically varies in resolution, accuracy, and signal-to-noise ratio, necessitating fusion techniques to enhance overall data quality. Implementation typically involves basic probability assignment functions and Dempster's combination rule, which mathematically merges evidence from different sources while managing conflicts through normalization factors. This paper presents practical case studies illustrating DS evidence theory applications in data fusion and methods to maximize its benefits. Key implementation considerations include handling conflicting evidence using algorithms like Murphy's average combination or more advanced conflict redistribution methods. Additionally, we provide detailed explanations and comprehensive references to help readers gain deeper understanding of the theory and its applications in the data fusion domain, including code structures for evidence combination functions and conflict measurement metrics.