Reflection and Action Under Scarce Resources: Theoretical Principles and Empirical Study

Eric Horvitz, Gregory Cooper, David Heckerman

Access pdf or postscript.

Abstract:

We define and exercise the expected value of computation as a fundamental component of reflection about alternative inference strategies. We present a portion of Protos research focused on the interlacing of reflection and action under scarce resources, and discuss how the techniques have been applied in a high-stakes medical domain. The work centers on endowing a computational agent with the ability to harness incomplete characterizations of problem-solving performance to control the amount of effort applied to a problem or subproblem, before taking action in the world or turning to another problem. We explore the use of the techniques in controlling decision-theoretic inference itself, and pose the approach as a model of rationality under scarce resources.

Keywords: Decision-theoretic metareasoning, bounded optimality, flexible computation, Bayesian networks, rationality under bounded resources, automated reasoning, influence diagrams, decision-theoretic inference.

Published in: Proceedings of the Eleventh International Joint Conference on Artificial Intelligence, Detroit, MI, pages 1121-1127. Morgan Kaufmann, San Mateo, CA, August 1989.. Also, Stanford CS Technical Report KSL-89-1.



Author Email: horvitz@microsoft.com



Related articles


Back to Eric Horvitz's home page.