Learning one subprocedure per lesson

Research output: Contribution to journalArticle

61 Citations (Scopus)

Abstract

sierra is a program that learns procedures incrementally from examples, where an example is a sequence of actions. sierra learns by completing explanations. Whenever the current procedure is inadequate for explaining (parsing) the current example, sierra formulates a new subprocedure whose instantiation completes the explanation (parse tree). The key to sierra's success lies in supplying a small amount of extra information with the examples. Instead of giving it a set of examples, under which conditions correct learning is provably impossible, it is given a sequence of "lessons," where a lesson is a set of examples that is guaranteed to introduce only one subprocedure. This permits unbiased learning, i.e., learning without a priori, heuristic preferences concerning the outcome.

Original languageEnglish (US)
Pages (from-to)1-40
Number of pages40
JournalArtificial Intelligence
Volume31
Issue number1
DOIs
StatePublished - 1987
Externally publishedYes

Fingerprint

learning
learning prerequisite
heuristics
Instantiation
Parsing
Heuristics

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computational Theory and Mathematics

Cite this

Learning one subprocedure per lesson. / VanLehn, Kurt.

In: Artificial Intelligence, Vol. 31, No. 1, 1987, p. 1-40.

Research output: Contribution to journalArticle

@article{6a81b2dbf76a42089359361342c02b88,
title = "Learning one subprocedure per lesson",
abstract = "sierra is a program that learns procedures incrementally from examples, where an example is a sequence of actions. sierra learns by completing explanations. Whenever the current procedure is inadequate for explaining (parsing) the current example, sierra formulates a new subprocedure whose instantiation completes the explanation (parse tree). The key to sierra's success lies in supplying a small amount of extra information with the examples. Instead of giving it a set of examples, under which conditions correct learning is provably impossible, it is given a sequence of {"}lessons,{"} where a lesson is a set of examples that is guaranteed to introduce only one subprocedure. This permits unbiased learning, i.e., learning without a priori, heuristic preferences concerning the outcome.",
author = "Kurt VanLehn",
year = "1987",
doi = "10.1016/0004-3702(87)90080-4",
language = "English (US)",
volume = "31",
pages = "1--40",
journal = "Artificial Intelligence",
issn = "0004-3702",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Learning one subprocedure per lesson

AU - VanLehn, Kurt

PY - 1987

Y1 - 1987

N2 - sierra is a program that learns procedures incrementally from examples, where an example is a sequence of actions. sierra learns by completing explanations. Whenever the current procedure is inadequate for explaining (parsing) the current example, sierra formulates a new subprocedure whose instantiation completes the explanation (parse tree). The key to sierra's success lies in supplying a small amount of extra information with the examples. Instead of giving it a set of examples, under which conditions correct learning is provably impossible, it is given a sequence of "lessons," where a lesson is a set of examples that is guaranteed to introduce only one subprocedure. This permits unbiased learning, i.e., learning without a priori, heuristic preferences concerning the outcome.

AB - sierra is a program that learns procedures incrementally from examples, where an example is a sequence of actions. sierra learns by completing explanations. Whenever the current procedure is inadequate for explaining (parsing) the current example, sierra formulates a new subprocedure whose instantiation completes the explanation (parse tree). The key to sierra's success lies in supplying a small amount of extra information with the examples. Instead of giving it a set of examples, under which conditions correct learning is provably impossible, it is given a sequence of "lessons," where a lesson is a set of examples that is guaranteed to introduce only one subprocedure. This permits unbiased learning, i.e., learning without a priori, heuristic preferences concerning the outcome.

UR - http://www.scopus.com/inward/record.url?scp=0023151233&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0023151233&partnerID=8YFLogxK

U2 - 10.1016/0004-3702(87)90080-4

DO - 10.1016/0004-3702(87)90080-4

M3 - Article

VL - 31

SP - 1

EP - 40

JO - Artificial Intelligence

JF - Artificial Intelligence

SN - 0004-3702

IS - 1

ER -