TY - JOUR
T1 - Epistemic Utility Estimation With Valuational Convexity
AU - Kafedziski, Venceslav
AU - Morrell, Darryl
N1 - Funding Information:
In classical Bayesian decision theory, it is assumed that the decision agent has a value system that can be expressed as a unique (up to a linear transform) utility function; the desirability of potential decisions are ranked according to this utility function, and decisions Manuscript received June 11, 1993; revised January 30, 1994. This work was supported by NSF grant MIP-9111174. The authors are with the Telecommunication Research Center and Department of Electrical Engineering, Arizona State University, Tempe, AZ 85287 USA. IEEE Log Number 9404539.
PY - 1995/1
Y1 - 1995/1
N2 - We address the problem of parameter estimation using Isaac Levi's epistemological framework of rational decision making; the resulting estimators accurately reflect the decision agent's knowledge, willingness to risk error, and values. In this paper, previous results are extended to include valuational convexity, yielding estimators that address conflicted values.
AB - We address the problem of parameter estimation using Isaac Levi's epistemological framework of rational decision making; the resulting estimators accurately reflect the decision agent's knowledge, willingness to risk error, and values. In this paper, previous results are extended to include valuational convexity, yielding estimators that address conflicted values.
KW - Epistemic utility estimation
KW - credal convexity
KW - set-valued estimation
KW - valuational convexity
UR - http://www.scopus.com/inward/record.url?scp=0029225284&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0029225284&partnerID=8YFLogxK
U2 - 10.1109/21.362953
DO - 10.1109/21.362953
M3 - Article
AN - SCOPUS:0029225284
SN - 0018-9472
VL - 25
SP - 190
EP - 193
JO - IEEE Transactions on Systems, Man, and Cybernetics
JF - IEEE Transactions on Systems, Man, and Cybernetics
IS - 1
ER -