Evidence Against Evidence Theory (?!)
This work tackles a foundational issue in uncertainty modeling for researchers in AI and statistics, offering an incremental improvement to MTE interpretations.
The paper addresses the weakness of the Mathematical Theory of Evidence (MTE) in relating to frequencies, demonstrating flaws in existing probabilistic interpretations and presenting a new frequency-based interpretation to overcome these drawbacks.
This paper is concerned with the apparent greatest weakness of the Mathematical Theory of Evidence (MTE) of Shafer \cite{Shafer:76}, which has been strongly criticized by Wasserman \cite{Wasserman:92ijar} - the relationship to frequencies. Weaknesses of various proposals of probabilistic interpretation of MTE belief functions are demonstrated. A new frequency-based interpretation is presented overcoming various drawbacks of earlier interpretations.