Premium
ASSESSMENT OF MULTIATTRIBUTED MEASURABLE VALUE AND UTILITY FUNCTIONS VIA MATHEMATICAL PROGRAMMING *
Author(s) -
Klein Gary,
Moskowitz Herbert,
Mahesh Sathiadev,
Ravindran A.
Publication year - 1985
Publication title -
decision sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.238
H-Index - 108
eISSN - 1540-5915
pISSN - 0011-7315
DOI - 10.1111/j.1540-5915.1985.tb01682.x
Subject(s) - computer science , consistency (knowledge bases) , function (biology) , decision maker , independence (probability theory) , mathematical optimization , value (mathematics) , process (computing) , expected utility hypothesis , scaling , operations research , artificial intelligence , machine learning , mathematical economics , mathematics , statistics , programming language , evolutionary biology , biology , geometry
A major restriction on the use of decision analysis in practice is the frequent difficulty of determining a decision maker's multiattribute utility function. The assessment process can be complex and tedious and generally involves: (1) identifying relevant independence conditions, (2) assessing conditional utility functions, (3) assessing scaling constants, and (4) checking for consistency. Some of the assessment and modeling complexities encountered include an assessor's inability to respond in a quantitatively meaningful and consistent way to hypothetical gambles and an analyst's problem in selecting an appropriate functional form that accurately characterizes the conditional utility assessments. A simplified procedure that mitigates these difficulties is proposed. This procedure facilitates the determination of scaling constants by obtaining (via mathematical programming) a multiattributed measurable value function which is converted to a multiattributed utility function. The methodology can be developed advantageously to produce an interactive software package for use as an assessment aid.