Strength or Accuracy: Credit Assignment in Learning Classifier Systems:
Classifier systems are an intriguing approach to a broad range of machine learning problems, based on automated generation and evaluation of condi tion/action rules. Inreinforcement learning tasks they simultaneously address the two major problems of learning a policy and generalising over it (and...
Gespeichert in:
| 1. Verfasser: | |
|---|---|
| Format: | Elektronisch E-Book |
| Sprache: | English |
| Veröffentlicht: |
London
Springer London
2004
|
| Ausgabe: | 1st ed. 2004 |
| Schriftenreihe: | Distinguished Dissertations
|
| Schlagworte: | |
| Online-Zugang: | UBY01 Volltext |
| Zusammenfassung: | Classifier systems are an intriguing approach to a broad range of machine learning problems, based on automated generation and evaluation of condi tion/action rules. Inreinforcement learning tasks they simultaneously address the two major problems of learning a policy and generalising over it (and re lated objects, such as value functions). Despite over 20 years of research, however, classifier systems have met with mixed success, for reasons which were often unclear. Finally, in 1995 Stewart Wilson claimed a long-awaited breakthrough with his XCS system, which differs from earlier classifier sys tems in a number of respects, the most significant of which is the way in which it calculates the value of rules for use by the rule generation system. Specifically, XCS (like most classifiersystems) employs a genetic algorithm for rule generation, and the way in whichit calculates rule fitness differsfrom earlier systems. Wilson described XCS as an accuracy-based classifiersystem and earlier systems as strength-based. The two differin that in strength-based systems the fitness of a rule is proportional to the return (reward/payoff) it receives, whereas in XCS it is a function of the accuracy with which return is predicted. The difference is thus one of credit assignment, that is, of how a rule's contribution to the system's performance is estimated. XCS is a Q learning system; in fact, it is a proper generalisation of tabular Q-learning, in which rules aggregate states and actions. In XCS, as in other Q-learners, Q-valuesare used to weightaction selection |
| Beschreibung: | 1 Online-Ressource (XVI, 307 p) |
| ISBN: | 9780857294166 |
| DOI: | 10.1007/978-0-85729-416-6 |
Internformat
MARC
| LEADER | 00000nmm a2200000zc 4500 | ||
|---|---|---|---|
| 001 | BV047064030 | ||
| 003 | DE-604 | ||
| 005 | 00000000000000.0 | ||
| 007 | cr|uuu---uuuuu | ||
| 008 | 201216s2004 |||| o||u| ||||||eng d | ||
| 020 | |a 9780857294166 |9 978-0-85729-416-6 | ||
| 024 | 7 | |a 10.1007/978-0-85729-416-6 |2 doi | |
| 035 | |a (ZDB-2-SCS)978-0-85729-416-6 | ||
| 035 | |a (OCoLC)1227479216 | ||
| 035 | |a (DE-599)BVBBV047064030 | ||
| 040 | |a DE-604 |b ger |e aacr | ||
| 041 | 0 | |a eng | |
| 049 | |a DE-706 | ||
| 082 | 0 | |a 006.3 |2 23 | |
| 100 | 1 | |a Kovacs, Tim |e Verfasser |4 aut | |
| 245 | 1 | 0 | |a Strength or Accuracy: Credit Assignment in Learning Classifier Systems |c by Tim Kovacs |
| 250 | |a 1st ed. 2004 | ||
| 264 | 1 | |a London |b Springer London |c 2004 | |
| 300 | |a 1 Online-Ressource (XVI, 307 p) | ||
| 336 | |b txt |2 rdacontent | ||
| 337 | |b c |2 rdamedia | ||
| 338 | |b cr |2 rdacarrier | ||
| 490 | 0 | |a Distinguished Dissertations | |
| 520 | |a Classifier systems are an intriguing approach to a broad range of machine learning problems, based on automated generation and evaluation of condi tion/action rules. Inreinforcement learning tasks they simultaneously address the two major problems of learning a policy and generalising over it (and re lated objects, such as value functions). Despite over 20 years of research, however, classifier systems have met with mixed success, for reasons which were often unclear. Finally, in 1995 Stewart Wilson claimed a long-awaited breakthrough with his XCS system, which differs from earlier classifier sys tems in a number of respects, the most significant of which is the way in which it calculates the value of rules for use by the rule generation system. Specifically, XCS (like most classifiersystems) employs a genetic algorithm for rule generation, and the way in whichit calculates rule fitness differsfrom earlier systems. Wilson described XCS as an accuracy-based classifiersystem and earlier systems as strength-based. The two differin that in strength-based systems the fitness of a rule is proportional to the return (reward/payoff) it receives, whereas in XCS it is a function of the accuracy with which return is predicted. The difference is thus one of credit assignment, that is, of how a rule's contribution to the system's performance is estimated. XCS is a Q learning system; in fact, it is a proper generalisation of tabular Q-learning, in which rules aggregate states and actions. In XCS, as in other Q-learners, Q-valuesare used to weightaction selection | ||
| 650 | 4 | |a Artificial Intelligence | |
| 650 | 4 | |a Algorithm Analysis and Problem Complexity | |
| 650 | 4 | |a Computer Appl. in Administrative Data Processing | |
| 650 | 4 | |a Artificial intelligence | |
| 650 | 4 | |a Algorithms | |
| 650 | 4 | |a Application software | |
| 655 | 7 | |0 (DE-588)4113937-9 |a Hochschulschrift |2 gnd-content | |
| 776 | 0 | 8 | |i Erscheint auch als |n Druck-Ausgabe |z 9781447110583 |
| 776 | 0 | 8 | |i Erscheint auch als |n Druck-Ausgabe |z 9781852337704 |
| 776 | 0 | 8 | |i Erscheint auch als |n Druck-Ausgabe |z 9780857294173 |
| 856 | 4 | 0 | |u https://doi.org/10.1007/978-0-85729-416-6 |x Verlag |z URL des Eerstveröffentlichers |3 Volltext |
| 912 | |a ZDB-2-SCS | ||
| 940 | 1 | |q ZDB-2-SCS_2000/2004 | |
| 999 | |a oai:aleph.bib-bvb.de:BVB01-032471142 | ||
| 966 | e | |u https://doi.org/10.1007/978-0-85729-416-6 |l UBY01 |p ZDB-2-SCS |q ZDB-2-SCS_2000/2004 |x Verlag |3 Volltext | |
Datensatz im Suchindex
| _version_ | 1804182061529956352 |
|---|---|
| adam_txt | |
| any_adam_object | |
| any_adam_object_boolean | |
| author | Kovacs, Tim |
| author_facet | Kovacs, Tim |
| author_role | aut |
| author_sort | Kovacs, Tim |
| author_variant | t k tk |
| building | Verbundindex |
| bvnumber | BV047064030 |
| collection | ZDB-2-SCS |
| ctrlnum | (ZDB-2-SCS)978-0-85729-416-6 (OCoLC)1227479216 (DE-599)BVBBV047064030 |
| dewey-full | 006.3 |
| dewey-hundreds | 000 - Computer science, information, general works |
| dewey-ones | 006 - Special computer methods |
| dewey-raw | 006.3 |
| dewey-search | 006.3 |
| dewey-sort | 16.3 |
| dewey-tens | 000 - Computer science, information, general works |
| discipline | Informatik |
| discipline_str_mv | Informatik |
| doi_str_mv | 10.1007/978-0-85729-416-6 |
| edition | 1st ed. 2004 |
| format | Electronic eBook |
| fullrecord | <?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>03363nmm a2200493zc 4500</leader><controlfield tag="001">BV047064030</controlfield><controlfield tag="003">DE-604</controlfield><controlfield tag="005">00000000000000.0</controlfield><controlfield tag="007">cr|uuu---uuuuu</controlfield><controlfield tag="008">201216s2004 |||| o||u| ||||||eng d</controlfield><datafield tag="020" ind1=" " ind2=" "><subfield code="a">9780857294166</subfield><subfield code="9">978-0-85729-416-6</subfield></datafield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/978-0-85729-416-6</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ZDB-2-SCS)978-0-85729-416-6</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OCoLC)1227479216</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)BVBBV047064030</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-604</subfield><subfield code="b">ger</subfield><subfield code="e">aacr</subfield></datafield><datafield tag="041" ind1="0" ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="049" ind1=" " ind2=" "><subfield code="a">DE-706</subfield></datafield><datafield tag="082" ind1="0" ind2=" "><subfield code="a">006.3</subfield><subfield code="2">23</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Kovacs, Tim</subfield><subfield code="e">Verfasser</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Strength or Accuracy: Credit Assignment in Learning Classifier Systems</subfield><subfield code="c">by Tim Kovacs</subfield></datafield><datafield tag="250" ind1=" " ind2=" "><subfield code="a">1st ed. 2004</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="a">London</subfield><subfield code="b">Springer London</subfield><subfield code="c">2004</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">1 Online-Ressource (XVI, 307 p)</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="490" ind1="0" ind2=" "><subfield code="a">Distinguished Dissertations</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Classifier systems are an intriguing approach to a broad range of machine learning problems, based on automated generation and evaluation of condi tion/action rules. Inreinforcement learning tasks they simultaneously address the two major problems of learning a policy and generalising over it (and re lated objects, such as value functions). Despite over 20 years of research, however, classifier systems have met with mixed success, for reasons which were often unclear. Finally, in 1995 Stewart Wilson claimed a long-awaited breakthrough with his XCS system, which differs from earlier classifier sys tems in a number of respects, the most significant of which is the way in which it calculates the value of rules for use by the rule generation system. Specifically, XCS (like most classifiersystems) employs a genetic algorithm for rule generation, and the way in whichit calculates rule fitness differsfrom earlier systems. Wilson described XCS as an accuracy-based classifiersystem and earlier systems as strength-based. The two differin that in strength-based systems the fitness of a rule is proportional to the return (reward/payoff) it receives, whereas in XCS it is a function of the accuracy with which return is predicted. The difference is thus one of credit assignment, that is, of how a rule's contribution to the system's performance is estimated. XCS is a Q learning system; in fact, it is a proper generalisation of tabular Q-learning, in which rules aggregate states and actions. In XCS, as in other Q-learners, Q-valuesare used to weightaction selection</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Artificial Intelligence</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Algorithm Analysis and Problem Complexity</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Computer Appl. in Administrative Data Processing</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Artificial intelligence</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Algorithms</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Application software</subfield></datafield><datafield tag="655" ind1=" " ind2="7"><subfield code="0">(DE-588)4113937-9</subfield><subfield code="a">Hochschulschrift</subfield><subfield code="2">gnd-content</subfield></datafield><datafield tag="776" ind1="0" ind2="8"><subfield code="i">Erscheint auch als</subfield><subfield code="n">Druck-Ausgabe</subfield><subfield code="z">9781447110583</subfield></datafield><datafield tag="776" ind1="0" ind2="8"><subfield code="i">Erscheint auch als</subfield><subfield code="n">Druck-Ausgabe</subfield><subfield code="z">9781852337704</subfield></datafield><datafield tag="776" ind1="0" ind2="8"><subfield code="i">Erscheint auch als</subfield><subfield code="n">Druck-Ausgabe</subfield><subfield code="z">9780857294173</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.1007/978-0-85729-416-6</subfield><subfield code="x">Verlag</subfield><subfield code="z">URL des Eerstveröffentlichers</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ZDB-2-SCS</subfield></datafield><datafield tag="940" ind1="1" ind2=" "><subfield code="q">ZDB-2-SCS_2000/2004</subfield></datafield><datafield tag="999" ind1=" " ind2=" "><subfield code="a">oai:aleph.bib-bvb.de:BVB01-032471142</subfield></datafield><datafield tag="966" ind1="e" ind2=" "><subfield code="u">https://doi.org/10.1007/978-0-85729-416-6</subfield><subfield code="l">UBY01</subfield><subfield code="p">ZDB-2-SCS</subfield><subfield code="q">ZDB-2-SCS_2000/2004</subfield><subfield code="x">Verlag</subfield><subfield code="3">Volltext</subfield></datafield></record></collection> |
| genre | (DE-588)4113937-9 Hochschulschrift gnd-content |
| genre_facet | Hochschulschrift |
| id | DE-604.BV047064030 |
| illustrated | Not Illustrated |
| index_date | 2024-07-03T16:12:21Z |
| indexdate | 2024-07-10T09:01:34Z |
| institution | BVB |
| isbn | 9780857294166 |
| language | English |
| oai_aleph_id | oai:aleph.bib-bvb.de:BVB01-032471142 |
| oclc_num | 1227479216 |
| open_access_boolean | |
| owner | DE-706 |
| owner_facet | DE-706 |
| physical | 1 Online-Ressource (XVI, 307 p) |
| psigel | ZDB-2-SCS ZDB-2-SCS_2000/2004 ZDB-2-SCS ZDB-2-SCS_2000/2004 |
| publishDate | 2004 |
| publishDateSearch | 2004 |
| publishDateSort | 2004 |
| publisher | Springer London |
| record_format | marc |
| series2 | Distinguished Dissertations |
| spelling | Kovacs, Tim Verfasser aut Strength or Accuracy: Credit Assignment in Learning Classifier Systems by Tim Kovacs 1st ed. 2004 London Springer London 2004 1 Online-Ressource (XVI, 307 p) txt rdacontent c rdamedia cr rdacarrier Distinguished Dissertations Classifier systems are an intriguing approach to a broad range of machine learning problems, based on automated generation and evaluation of condi tion/action rules. Inreinforcement learning tasks they simultaneously address the two major problems of learning a policy and generalising over it (and re lated objects, such as value functions). Despite over 20 years of research, however, classifier systems have met with mixed success, for reasons which were often unclear. Finally, in 1995 Stewart Wilson claimed a long-awaited breakthrough with his XCS system, which differs from earlier classifier sys tems in a number of respects, the most significant of which is the way in which it calculates the value of rules for use by the rule generation system. Specifically, XCS (like most classifiersystems) employs a genetic algorithm for rule generation, and the way in whichit calculates rule fitness differsfrom earlier systems. Wilson described XCS as an accuracy-based classifiersystem and earlier systems as strength-based. The two differin that in strength-based systems the fitness of a rule is proportional to the return (reward/payoff) it receives, whereas in XCS it is a function of the accuracy with which return is predicted. The difference is thus one of credit assignment, that is, of how a rule's contribution to the system's performance is estimated. XCS is a Q learning system; in fact, it is a proper generalisation of tabular Q-learning, in which rules aggregate states and actions. In XCS, as in other Q-learners, Q-valuesare used to weightaction selection Artificial Intelligence Algorithm Analysis and Problem Complexity Computer Appl. in Administrative Data Processing Artificial intelligence Algorithms Application software (DE-588)4113937-9 Hochschulschrift gnd-content Erscheint auch als Druck-Ausgabe 9781447110583 Erscheint auch als Druck-Ausgabe 9781852337704 Erscheint auch als Druck-Ausgabe 9780857294173 https://doi.org/10.1007/978-0-85729-416-6 Verlag URL des Eerstveröffentlichers Volltext |
| spellingShingle | Kovacs, Tim Strength or Accuracy: Credit Assignment in Learning Classifier Systems Artificial Intelligence Algorithm Analysis and Problem Complexity Computer Appl. in Administrative Data Processing Artificial intelligence Algorithms Application software |
| subject_GND | (DE-588)4113937-9 |
| title | Strength or Accuracy: Credit Assignment in Learning Classifier Systems |
| title_auth | Strength or Accuracy: Credit Assignment in Learning Classifier Systems |
| title_exact_search | Strength or Accuracy: Credit Assignment in Learning Classifier Systems |
| title_exact_search_txtP | Strength or Accuracy: Credit Assignment in Learning Classifier Systems |
| title_full | Strength or Accuracy: Credit Assignment in Learning Classifier Systems by Tim Kovacs |
| title_fullStr | Strength or Accuracy: Credit Assignment in Learning Classifier Systems by Tim Kovacs |
| title_full_unstemmed | Strength or Accuracy: Credit Assignment in Learning Classifier Systems by Tim Kovacs |
| title_short | Strength or Accuracy: Credit Assignment in Learning Classifier Systems |
| title_sort | strength or accuracy credit assignment in learning classifier systems |
| topic | Artificial Intelligence Algorithm Analysis and Problem Complexity Computer Appl. in Administrative Data Processing Artificial intelligence Algorithms Application software |
| topic_facet | Artificial Intelligence Algorithm Analysis and Problem Complexity Computer Appl. in Administrative Data Processing Artificial intelligence Algorithms Application software Hochschulschrift |
| url | https://doi.org/10.1007/978-0-85729-416-6 |
| work_keys_str_mv | AT kovacstim strengthoraccuracycreditassignmentinlearningclassifiersystems |