An Analysis of Categorical and Quantitative Methods for Planning under Uncertainty

From Tetherless World Wiki

Jump to: navigation, search

Citation: Curtis Langlotz and Edward H. Shortliffe. (1988) An Analysis of Categorical and Quantitative Methods for Planning under Uncertainty. In KSL-88-63, 1988.

Publication techreport ( Edit )
type Technical Report
bibtype techreport
Bibtex basics
author Curtis Langlotz and Edward H. Shortliffe
title An Analysis of Categorical and Quantitative Methods for Planning under Uncertainty
number KSL-88-63
institution Knowledge Systems, AI Laboratory
year 1988
Bibtex more
Access Paper
abstract Decision theory and logical reasoning are both methods for representing and solving medical decision problems. We analyze the usefulness of these two approaches to medical therapy planning by establishing a simple correspondence between decision theory and non-monotonic logic, a formalization of categorical logical reasoning. The analysis indicates that categorical approaches to planning can be viewed as comprising two decision-theoretic concepts: probabilities (degrees of belief in planning hypotheses) and utilities (degrees of desirability of planning outcomes). We present and discuss examples of the following lessons from this decision-theoretic view of categorical (nonmonotonic) reasoning: (1) Decision theory and artificial intelligence techniques are intended to solve different components of the planning problem. (2) When considered in the context of planning under uncertainty, nonmonotonic logics do not retain the domain-independent characteristics of classical logical reasoning for planning under certainty. (3) Because certain nonmonotonic programming paradigms (e.g., frame-based inheritance, rule-based planning, protocol-based reminders) are inherently problem-specific, they may be inappropriate to employ in the solution of certain types of planning problems. We discuss how these conclusions affect several current medical informatics research issues, including the construction of "very large" medical knowledge bases.

KSL Technical Report ID: KSL-88-63
Facts about An Analysis of Categorical and Quantitative Methods for Planning under UncertaintyRDF feed
Abstract Decision theory and logical reasoning are … Decision theory and logical reasoning are both methods for representing and solving medical decision problems. We analyze the usefulness of these two approaches to medical therapy planning by establishing a simple correspondence between decision theory and non-monotonic logic, a formalization of categorical logical reasoning. The analysis indicates that categorical approaches to planning can be viewed as comprising two decision-theoretic concepts: probabilities (degrees of belief in planning hypotheses) and utilities (degrees of desirability of planning outcomes). We present and discuss examples of the following lessons from this decision-theoretic view of categorical (nonmonotonic) reasoning: (1) Decision theory and artificial intelligence techniques are intended to solve different components of the planning problem. (2) When considered in the context of planning under uncertainty, nonmonotonic logics do not retain the domain-independent characteristics of classical logical reasoning for planning under certainty. (3) Because certain nonmonotonic programming paradigms (e.g., frame-based inheritance, rule-based planning, protocol-based reminders) are inherently problem-specific, they may be inappropriate to employ in the solution of certain types of planning problems. We discuss how these conclusions affect several current medical informatics research issues, including the construction of "very large" medical knowledge bases. n of "very large" medical knowledge bases.
Author Curtis Langlotz and Edward H. Shortliffe  +
Bibtype techreport  +
Has author Curtis Langlotz and Edward H. Shortliffe  +
Has identifier KSL-88-63  +
Has publishing details 1988  +
Has title An Analysis of Categorical and Quantitative Methods for Planning under Uncertainty  +
Has where published KSL-88-63  +
Has year 1988  +
Institution Knowledge Systems, AI Laboratory  +
Ksl tr id KSL-88-63  +
Number KSL-88-63  +
Process note YES  +
Title An Analysis of Categorical and Quantitative Methods for Planning under Uncertainty  +
Year 1988  +
Retrieved from "http://tw.rpi.edu/wiki/An_Analysis_of_Categorical_and_Quantitative_Methods_for_Planning_under_Uncertainty"
Personal tools