Automatic Discovery of Heuristics for Nondeterministic Programs from Sample Execution Traces

During the last few years a number of relatively effective Artificial Intelligence (AI) programs have been written incorporating considerable amounts of problem specific knowledge. In particular, declarative representations have attracted much attention partly because of the relative ease with which...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Stolfo,Salvatore J
Format: Report
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Stolfo,Salvatore J
description During the last few years a number of relatively effective Artificial Intelligence (AI) programs have been written incorporating considerable amounts of problem specific knowledge. In particular, declarative representations have attracted much attention partly because of the relative ease with which knowledge can be communicated in this form. Unfortunately, straightforward implementation of declaratively specified knowledge corresponds to a nondeterministic program which incurs enormous computational costs. This thesis investigates one way to limit this cost. We develop control heuristics for a family of problems from traces of sample solutions generated during a training session with a human expert. Algorithms are presented which recognize a set of patterns in the sequence of knowledge applications and which compile descriptions of these patterns in a control language, called CRAPS. More specifically, patterns of repeating, parallel and common sequences are considered in the analysis. The analysis also produces a set of meta-rules which aid the CRAPS description in the event the sequencing it specifies is inappropriate. The CRAPS description and meta-rules are then used for guidance in solving subsequent problems.
format Report
fullrecord <record><control><sourceid>dtic_1RU</sourceid><recordid>TN_cdi_dtic_stinet_ADA076767</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>ADA076767</sourcerecordid><originalsourceid>FETCH-dtic_stinet_ADA0767673</originalsourceid><addsrcrecordid>eNqFjDEKAjEQRdNYiHoDi7mAIAhaB3dlKxHcUlhCdiKBTUZmEtHbO4i9_OLDe_Dm5mZroeRK9NBE8fREfgMF6LByFMUCgRjOlEcsyCnmL4UL051dUsuU4OrSY0JoX-hriZShZ-dRlmYW3CS4-v3CrE9tf-w2oyYG7WQsg23s9rDX7f7oD5kKOXU</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>report</recordtype></control><display><type>report</type><title>Automatic Discovery of Heuristics for Nondeterministic Programs from Sample Execution Traces</title><source>DTIC Technical Reports</source><creator>Stolfo,Salvatore J</creator><creatorcontrib>Stolfo,Salvatore J ; NEW YORK UNIV N Y COURANT INST OF MATHEMATICAL SCIENCES</creatorcontrib><description>During the last few years a number of relatively effective Artificial Intelligence (AI) programs have been written incorporating considerable amounts of problem specific knowledge. In particular, declarative representations have attracted much attention partly because of the relative ease with which knowledge can be communicated in this form. Unfortunately, straightforward implementation of declaratively specified knowledge corresponds to a nondeterministic program which incurs enormous computational costs. This thesis investigates one way to limit this cost. We develop control heuristics for a family of problems from traces of sample solutions generated during a training session with a human expert. Algorithms are presented which recognize a set of patterns in the sequence of knowledge applications and which compile descriptions of these patterns in a control language, called CRAPS. More specifically, patterns of repeating, parallel and common sequences are considered in the analysis. The analysis also produces a set of meta-rules which aid the CRAPS description in the event the sequencing it specifies is inappropriate. The CRAPS description and meta-rules are then used for guidance in solving subsequent problems.</description><language>eng</language><subject>ALGORITHMS ; ARTIFICIAL INTELLIGENCE ; CODING ; Computer Programming and Software ; COMPUTER PROGRAMS ; CONTROL SYSTEMS ; Cybernetics ; HEURISTIC METHODS ; OPERATORS(MATHEMATICS) ; PATTERN RECOGNITION ; PROGRAMMING LANGUAGES</subject><creationdate>1979</creationdate><rights>APPROVED FOR PUBLIC RELEASE</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,780,885,27567,27568</link.rule.ids><linktorsrc>$$Uhttps://apps.dtic.mil/sti/citations/ADA076767$$EView_record_in_DTIC$$FView_record_in_$$GDTIC$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>Stolfo,Salvatore J</creatorcontrib><creatorcontrib>NEW YORK UNIV N Y COURANT INST OF MATHEMATICAL SCIENCES</creatorcontrib><title>Automatic Discovery of Heuristics for Nondeterministic Programs from Sample Execution Traces</title><description>During the last few years a number of relatively effective Artificial Intelligence (AI) programs have been written incorporating considerable amounts of problem specific knowledge. In particular, declarative representations have attracted much attention partly because of the relative ease with which knowledge can be communicated in this form. Unfortunately, straightforward implementation of declaratively specified knowledge corresponds to a nondeterministic program which incurs enormous computational costs. This thesis investigates one way to limit this cost. We develop control heuristics for a family of problems from traces of sample solutions generated during a training session with a human expert. Algorithms are presented which recognize a set of patterns in the sequence of knowledge applications and which compile descriptions of these patterns in a control language, called CRAPS. More specifically, patterns of repeating, parallel and common sequences are considered in the analysis. The analysis also produces a set of meta-rules which aid the CRAPS description in the event the sequencing it specifies is inappropriate. The CRAPS description and meta-rules are then used for guidance in solving subsequent problems.</description><subject>ALGORITHMS</subject><subject>ARTIFICIAL INTELLIGENCE</subject><subject>CODING</subject><subject>Computer Programming and Software</subject><subject>COMPUTER PROGRAMS</subject><subject>CONTROL SYSTEMS</subject><subject>Cybernetics</subject><subject>HEURISTIC METHODS</subject><subject>OPERATORS(MATHEMATICS)</subject><subject>PATTERN RECOGNITION</subject><subject>PROGRAMMING LANGUAGES</subject><fulltext>true</fulltext><rsrctype>report</rsrctype><creationdate>1979</creationdate><recordtype>report</recordtype><sourceid>1RU</sourceid><recordid>eNqFjDEKAjEQRdNYiHoDi7mAIAhaB3dlKxHcUlhCdiKBTUZmEtHbO4i9_OLDe_Dm5mZroeRK9NBE8fREfgMF6LByFMUCgRjOlEcsyCnmL4UL051dUsuU4OrSY0JoX-hriZShZ-dRlmYW3CS4-v3CrE9tf-w2oyYG7WQsg23s9rDX7f7oD5kKOXU</recordid><startdate>197909</startdate><enddate>197909</enddate><creator>Stolfo,Salvatore J</creator><scope>1RU</scope><scope>BHM</scope></search><sort><creationdate>197909</creationdate><title>Automatic Discovery of Heuristics for Nondeterministic Programs from Sample Execution Traces</title><author>Stolfo,Salvatore J</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-dtic_stinet_ADA0767673</frbrgroupid><rsrctype>reports</rsrctype><prefilter>reports</prefilter><language>eng</language><creationdate>1979</creationdate><topic>ALGORITHMS</topic><topic>ARTIFICIAL INTELLIGENCE</topic><topic>CODING</topic><topic>Computer Programming and Software</topic><topic>COMPUTER PROGRAMS</topic><topic>CONTROL SYSTEMS</topic><topic>Cybernetics</topic><topic>HEURISTIC METHODS</topic><topic>OPERATORS(MATHEMATICS)</topic><topic>PATTERN RECOGNITION</topic><topic>PROGRAMMING LANGUAGES</topic><toplevel>online_resources</toplevel><creatorcontrib>Stolfo,Salvatore J</creatorcontrib><creatorcontrib>NEW YORK UNIV N Y COURANT INST OF MATHEMATICAL SCIENCES</creatorcontrib><collection>DTIC Technical Reports</collection><collection>DTIC STINET</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Stolfo,Salvatore J</au><aucorp>NEW YORK UNIV N Y COURANT INST OF MATHEMATICAL SCIENCES</aucorp><format>book</format><genre>unknown</genre><ristype>RPRT</ristype><btitle>Automatic Discovery of Heuristics for Nondeterministic Programs from Sample Execution Traces</btitle><date>1979-09</date><risdate>1979</risdate><abstract>During the last few years a number of relatively effective Artificial Intelligence (AI) programs have been written incorporating considerable amounts of problem specific knowledge. In particular, declarative representations have attracted much attention partly because of the relative ease with which knowledge can be communicated in this form. Unfortunately, straightforward implementation of declaratively specified knowledge corresponds to a nondeterministic program which incurs enormous computational costs. This thesis investigates one way to limit this cost. We develop control heuristics for a family of problems from traces of sample solutions generated during a training session with a human expert. Algorithms are presented which recognize a set of patterns in the sequence of knowledge applications and which compile descriptions of these patterns in a control language, called CRAPS. More specifically, patterns of repeating, parallel and common sequences are considered in the analysis. The analysis also produces a set of meta-rules which aid the CRAPS description in the event the sequencing it specifies is inappropriate. The CRAPS description and meta-rules are then used for guidance in solving subsequent problems.</abstract><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier
ispartof
issn
language eng
recordid cdi_dtic_stinet_ADA076767
source DTIC Technical Reports
subjects ALGORITHMS
ARTIFICIAL INTELLIGENCE
CODING
Computer Programming and Software
COMPUTER PROGRAMS
CONTROL SYSTEMS
Cybernetics
HEURISTIC METHODS
OPERATORS(MATHEMATICS)
PATTERN RECOGNITION
PROGRAMMING LANGUAGES
title Automatic Discovery of Heuristics for Nondeterministic Programs from Sample Execution Traces
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T18%3A47%3A54IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-dtic_1RU&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=unknown&rft.btitle=Automatic%20Discovery%20of%20Heuristics%20for%20Nondeterministic%20Programs%20from%20Sample%20Execution%20Traces&rft.au=Stolfo,Salvatore%20J&rft.aucorp=NEW%20YORK%20UNIV%20N%20Y%20COURANT%20INST%20OF%20MATHEMATICAL%20SCIENCES&rft.date=1979-09&rft_id=info:doi/&rft_dat=%3Cdtic_1RU%3EADA076767%3C/dtic_1RU%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true