{"type":"ScholarlyArticle","title":"On Optimal Game-Tree Search using Rational Meta-Reasoning","authors":[{"name":"Stuart J. Russell","slug":"stuart-j-russell","given":"Stuart J.","family":"Russell","affiliation":"Computer Science Division University of California Berkeley , CA 94720"},{"name":"Eric Wefald","slug":"eric-wefald","given":"Eric","family":"Wefald","affiliation":"Computer Science Division University of California Berkeley , CA 94720"}],"abstract":"In this paper we outline a general approach to the study of problem-solving, in which search steps are considered decisions in the same sense as actions in the world. Unlike other metrics in the literature, the value of a search step is defined as a real utility rather than as a quasiutility, and can therefore be computed directly from a model of the base-level problem-solver. We develop a formula for the expected value of a search step in a game-playing context using the single-step assumption, namely that a computation step can be evaluated as it was the last to be taken. We prove some metalevel theorems that enable the development of a low-overhead algorithm, MGSS*, that chooses search steps in order of highest estimated utility. Although we show that the single-step assumption is untenable in general, a program implemented for the game of Othello soundly beats an alpha-beta search while expanding significantly fewer nodes, even though both programs use the same evaluation function.","doi":null,"url":"https://ijcai.org/proceedings/1989/53","pdf":"https://ijcai.org/proceedings/1989/0053.pdf","markdown":"https://ijcai.org/proceedings/1989/53.md","bibtex":"https://ijcai.org/proceedings/1989/bibtex/53","volume":{"slug":"1989","title":"Proceedings of the Eleventh International Joint Conference on Artificial Intelligence","short_title":"IJCAI 1989","url":"https://ijcai.org/proceedings/1989/","year":1989,"location":"Detroit, Michigan, USA","dates":"August 20-25 1989","editors":"N. S. Sridharan","isbn":"","publisher":"International Joint Conferences on Artificial Intelligence","doi_prefix":""},"section":"Search","track":"Search","number":53,"pages":{"first":334,"last":339},"keywords":[],"videos":[],"references":[{"position":1,"raw":"[Berliner, 1979] Berliner, .H. J. (1979) The B* Tr~e ~earch Algorithm: A Best-Fust Proof Procedure. Artlficaal In- telligence 12.","title":"The B* Tr~e ~earch Algorithm: A Best-Fust Proof Procedure","authors":". H J Berliner","year":"1979","venue":"Artlficaal Intelligence","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Berliner, 1980] Berliner, H. (1980). Backgammon computer program beats world champion. Artificial Intelligence, 14, 205-220.","title":"Backgammon computer program beats world champion","authors":"H Berliner","year":"1980","venue":"Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Campbell, 1988] Campbell, M. (1988) Singular Extensions: Adding Selectivity to Brute Force Searching. In Pro- ceedings of the AAAI Symposium on Computer Game- Playing, Stanford, CA, 8-13.","title":"Singular Extensions: Adding Selectivity to Brute Force Searching","authors":"M Campbell","year":"1988","venue":"Proceedings of the AAAI Symposium on Computer Game-Playing","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Conroy, forthcoming] Conroy, J. M. (forthcoming) Decision-theoretic control of search in probabilistic do- mains. MS Report, Computer Science Division, Univer- sity of California, Berkeley, CA.","title":"forthcoming) Decision-theoretic control of search in probabilistic domains","authors":"J M Conroy","year":null,"venue":null,"doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Good, 1968] Good, l J. (1968) A five year plan for auto-- matic chess. Machine Intelligence, 2.","title":"A five year plan for auto-matic chess","authors":"J Good ; Good","year":"1968","venue":"Machine Intelligence","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Good, 1977] Good, I. J. (1977) Dynamic probability, com- puter chess and the measurement of knowledge. Ma- chine Intelligence, 8.","title":"Dynamic probability, computer chess and the measurement of knowledge","authors":"I J Good","year":"1977","venue":"Machine Intelligence","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Howard, 1966] Howard, R. A. (1966) Information value the- ory. IEEE Transactions on Systems Science and Cyber- netics, SSC-2(1), 22-26.","title":"Information value theory","authors":"R A Howard","year":"1966","venue":"IEEE Transactions on Systems Science and Cybernetics","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[McAllester, 1988] McAllester, D.A. (1988) Conspiracy Numbers for Min-Max -Search. Artificial Intelligence, 35, 287-310.","title":"Conspiracy Numbers for Min-Max -Search","authors":"D A Mcallester","year":"1988","venue":"Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Palay, 1982] Palay, A. J. (1982) The B* Tree Search Algorithm-New Results. Artificial Intelligence 19.","title":"The B* Tree Search Algorithm-New Results","authors":"A J Palay","year":"1982","venue":"Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Palay, 1985] Palay, A.J. (1985) Searching with Probabilities. Marshfield, MA: Pitman :Publishing Inc. [Pearl, 1988]","title":"Searching with Probabilities","authors":"A J Palay","year":"1985","venue":null,"doi":null,"url":null,"internal_url":null},{"position":11,"raw":"Pearl, J. (1988) Probabilistic Reasoning in In- telligent Systems: Networks of Probable Inference San Mateo, CA: Morgan Kaufmann. [Raiffa and Schlaifer, 1961]","title":"Probabilistic Reasoning in Intelligent Systems: Networks of Probable Inference San Mateo","authors":"J Pearl","year":"1961","venue":"Raiffa and Schlaifer","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"Raiffa, H, and Schlaifer, R. (1961) Applied Statistical Decision Theory. Cambridge, MA: Harvard University Press.","title":"Applied Statistical Decision Theory","authors":"H Raiffa, R Schlaifer","year":"1961","venue":null,"doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Rivest, 1988] Rivest, R.L. (1988) Game Tree Searching by Min/Max Approximation. Artificial Intelligence 34.","title":"Game Tree Searching by Min/Max Approximation","authors":"R L Rivest","year":"1988","venue":"Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Russell and Wefald, 1988a] Russell, S. J., and Wefald, E. H. (1988a) Multi-Level Decision-Theoretic Search. Pro- ceedings of the AAAI Symposium on Computer Game- Playing, Stanford, CA, 3-7.","title":"Multi-Level Decision-Theoretic Search","authors":"Wefald ; Russell, S J Russell, E H Wefald","year":"1988","venue":"Proceedings of the AAAI Symposium on Computer Game-Playing","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Russell and Wefald, 1988b] Russell, S. J., and Wefald, E. H. (1988b) Decision-theoretic control of search: General theory and an application to game-playing. Technical report UCB/CSD 88/435, Computer Science Division, University of California, Berkeley, CA.","title":"Decision-theoretic control of search: General theory and an application to game-playing","authors":"Wefald ; Russell, S J Russell, E H Wefald","year":"1988","venue":null,"doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Russell and Wefald, 1989] Russell, S. J ., and Wefald, E. H. (1989) Principles of Metareasoning. In Proceedings of the First International Conference on Principles of Knowledge Representation and Reasoning, Toronto, On- tario: Morgan Kaufmann.","title":"Principles of Metareasoning","authors":"S J Russell, E H Wefald","year":"1989","venue":"Proceedings of the First International Conference on Principles of Knowledge Representation and Reasoning","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Wefald and Russell, 1989a] Wefald, E. H., and Russell, S. J . (1989a) Estimating the Value of Computation: the Case of Real-time Search. AAAI Spring Symposium on AI and Limited Rationality Working Notes Stanford, CA, March, 1989.","title":"Estimating the Value of Computation: the Case of Real-time Search","authors":"Russell ; Wefald, E H Wefald, S J Russell","year":"1989","venue":"AAAI Spring Symposium on AI and Limited Rationality Working Notes Stanford","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Wefald and Russell, 1989b] Wefald, E. H., and Russell, S. J. (1989b) Adaptive Learning of Decision-Theoretic Search Control Knowledge. Sixth International Work- shop on Machine Learning, Ithaca, NY, June, 1989.","title":"Adaptive Learning of Decision-Theoretic Search Control Knowledge","authors":"Russell ; Wefald, E H Wefald, S J Russell","year":"1989","venue":"Sixth International Workshop on Machine Learning","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}