Terminating Decision Algorithms Optimally

13 years 10 months ago

Download www.cs.cmu.edu

Incomplete decision algorithms can often solve larger problem instances than complete ones. The drawback is that one does not know whether the algorithm will ﬁnish soon, later, or never. This paper presents a general decision-theoretic method for optimally terminating such algorithms. The stopping policy is computed based on a prior probability of the answer, a payoff model describing the value that different probability estimates would provide at different times, and the algorithm’s runtime distribution. We present a linear-time algorithm for determining the optimal stopping policy given a ﬁnite cap on the number of algorithm steps. We exemplify this in a manufacturing scenario with a 3-satisﬁability problem. To increase accuracy, the initial satisﬁability probability and the run-time distribution are conditioned on features of the instance. The expectation of the result at each future time step is computed using Bayesian updating. We then extend the framework to settings w...

Tuomas Sandholm

Real-time Traffic