A Mixed Radix Approach to the Pooling of Evidence

Bell, David A.; Webb, Jerry; Guan, Jingqi

doi:10.1515/jisys.1995.5.1.1

Cited by 2 publications

(2 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using the tableau in Appendix B would be very difficult -although it is possible to implement the result tables in hardware and do a "flash" table look up (Bell et al 1994). Moreover it has the other deficiencies outlined above -mainly that it is not symmetrical and easy to remember and use.…”

Section: A Justification Of Our Simple Numeric Scale For Straight Decmentioning

confidence: 99%

“…Most traditional approaches to evidential reasoning are based on numerical methods of representing evidential supports (Shafer 1976;Bell 1991, 1993;Bell et al 1994;Yager et al 1994), but there have also been several studies on non-numerical and logical methods (Cohen 1985;An et al 1992aAn et al , 1993aBell 1993). Our starting point for the present study is a non-numerical method called RES, based on a logic for uncertain information representation and uncertain reasoning (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Graded Relative Evidence

Bell

2005

Artif Intell Rev

View full text Add to dashboard Cite

Relative Evidential Supports (RES) was developed and justified several years ago as a non-numeric apparatus that allows us to compare evidential supports for alternative conclusions when making a decision. An extension, called Graded Relative Evidence (GRE), of the RES concept of pairwise balancing and trading-off of evidence is reported here which keeps its basic features of simplicity and perspicacity, but enriches its modelling fidelity by permitting very modest, and intuitive, variations in degrees of outweighing (which the essentially binary RES does not). The formal justification is very simply based on linkages to RES and to the Dempster-Shafer theory of evidence. The use of the simple extension is illustrated, and to a small degree further justified empirically, by application to a topical scientific debate, about what is called the Congo Crossover Conjecture here. This decision-making instance is chosen because of the wealth of evidence that has been accumulated on both sides of the debate, and the range of evidence strengths manifested in it. The conjecture is that the advent of Aids was in the late 1950s in the Congo, when a vaccine for polio was allegedly cultivated in the kidneys of chimpanzees, which allowed the Aids infection to cross over to humans from primates.

show abstract

Section: A Justification Of Our Simple Numeric Scale For Straight Decmentioning

confidence: 99%