Skip to main content

2021 | OriginalPaper | Buchkapitel

On the Effects of Absumption for XCS with Continuous-Valued Inputs

verfasst von : Alexander R. M. Wagner, Anthony Stein

Erschienen in: Applications of Evolutionary Computation

Verlag: Springer International Publishing

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

The rule-based XCS Classifier System (XCS) aims at forming classifiers which are as general as possible to achieve an optimal performance level. A too high generalization pressure may lead to over-general classifiers degrading the performance of XCS. To date, no method exists for XCS for real-valued input spaces (XCSR) to handle over-general classifiers ensuring an accurate population. The Absumption mechanism and the Specify operator, both developed for XCS with binary inputs, provide a promising basis for over-generality handling in XCSR. This paper introduces adapted versions of Absumption and Specify by proposing different identification and specialization strategies for the application in XCSR. To determine their potential, the adapted techniques will be evaluated in different classification problems, i.e., common benchmarks and real-world data from the agricultural domain, and in a multi-step problem.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Fußnoten
1
6-RMP: N = 800, \(\alpha \) = 0.1, \(\beta \) = 0.2, \(\delta \) = 0.1, \(\nu \) = 5, \(\theta _{mna}\) = 2, \(\theta _{GA}\) = 12, \(\theta _{del}\) = 20, \(\theta _{sub}\) = 20, \(\epsilon _0\) = 10, \(\chi \) = 0.8, \(\mu \) = 0.04, \(p_{ini}\) = 10.0, \(\epsilon _{ini}\) = 0.0, \(F_{ini}\) = 0.01, \(\epsilon _{red}\) = 0.25, \(F_{red}\) = 0.1, \(m_0\) = 0.1, \(r_0\) = 1.0.
 
2
CBP(3,3): Analogous to 6-RMP, except: N = 2000, \(r_0\) = 0.5.
 
3
Mario: Analogous to 6-RMP, except: N = 7000, \(\beta \) = 0.3, \(\theta _{mna}\) = 6, \(\theta _{GA}\) = 30, \(\theta _{del}\) = 50, \(\theta _{sub}\) = 50, \(r_0\) = 0.1.
 
7
Analogous to 6-RMP, except: N = 6400, \(\theta _{GA}\) = 48, \(\theta _{del}\) = 50, \(\theta _{sub}\) = 50, \(\epsilon _0\) = 1.0, \(\epsilon _{red}\) = 1.0, \(m_0\) = 0.5, \(r_0\) = 1.0.
 
8
Analogous to 6-RMP, except: N = 10,000, \(\gamma \) = 0.95, \(\theta _{mna}\) = 4, \(\theta _{GA}\) = 50, \(\theta _{del}\) = 50, \(\theta _{sub}\) = 50, \(\epsilon _0\) = 0.005, \(m_0\) = 0.25, \(r_0\) = 0.5.
 
Literatur
1.
Zurück zum Zitat Bacardit, J., Burke, E.K., Krasnogor, N.: Improving the scalability of rule-based evolutionary learning. Memet. Comput. 1(1), 55–67 (2009)CrossRef Bacardit, J., Burke, E.K., Krasnogor, N.: Improving the scalability of rule-based evolutionary learning. Memet. Comput. 1(1), 55–67 (2009)CrossRef
2.
Zurück zum Zitat Barry, A.M.: The stability of long action chains in XCS. Soft Comput. 6(3), 183–199 (2002)CrossRef Barry, A.M.: The stability of long action chains in XCS. Soft Comput. 6(3), 183–199 (2002)CrossRef
3.
Zurück zum Zitat Bernadó-Mansilla, E., Garrell-Guiu, J.M.: Accuracy-based learning classifier systems: models, analysis and applications to classification tasks. Evol. Comput. 11(3), 209–238 (2003)CrossRef Bernadó-Mansilla, E., Garrell-Guiu, J.M.: Accuracy-based learning classifier systems: models, analysis and applications to classification tasks. Evol. Comput. 11(3), 209–238 (2003)CrossRef
4.
Zurück zum Zitat Butz, M.V., Kovacs, T., Lanzi, P.L., Wilson, S.W.: How XCS evolves accurate classifiers. In: Proceedings of GECCO 2001. GECCO 2001, pp. 927–934. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA (2001) Butz, M.V., Kovacs, T., Lanzi, P.L., Wilson, S.W.: How XCS evolves accurate classifiers. In: Proceedings of GECCO 2001. GECCO 2001, pp. 927–934. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA (2001)
5.
Zurück zum Zitat Butz, M.V., Lanzi, P.L., Wilson, S.W.: Function approximation with XCS: hyperellipsoidal conditions, recursive least squares, and compaction. IEEE T Evolut. Comput. 12(3), 355–376 (2008)CrossRef Butz, M.V., Lanzi, P.L., Wilson, S.W.: Function approximation with XCS: hyperellipsoidal conditions, recursive least squares, and compaction. IEEE T Evolut. Comput. 12(3), 355–376 (2008)CrossRef
8.
Zurück zum Zitat Calian, D.A., Bacardit, J.: Integrating memetic search into the BioHEL evolutionary learning system for large-scale datasets. Memet. Comput. 5(2), 95–130 (2013)CrossRef Calian, D.A., Bacardit, J.: Integrating memetic search into the BioHEL evolutionary learning system for large-scale datasets. Memet. Comput. 5(2), 95–130 (2013)CrossRef
10.
Zurück zum Zitat Holland, J.H., Reitman, J.S.: Cognitive systems based on adaptive algorithms. In: Pattern-Directed Inference Systems, pp. 313–329. Academic Press (1978) Holland, J.H., Reitman, J.S.: Cognitive systems based on adaptive algorithms. In: Pattern-Directed Inference Systems, pp. 313–329. Academic Press (1978)
11.
Zurück zum Zitat Kovacs, T.: Towards a theory of strong overgeneral classifiers. In: FOGA 6, pp. 165–184. Morgan Kaufmann, San Francisco (2001) Kovacs, T.: Towards a theory of strong overgeneral classifiers. In: FOGA 6, pp. 165–184. Morgan Kaufmann, San Francisco (2001)
12.
Zurück zum Zitat Lanzi, P.L.: A study on the generalization capabilities of XCS. In: Proceedings of ICGA 1997, pp. 418–425 (1997) Lanzi, P.L.: A study on the generalization capabilities of XCS. In: Proceedings of ICGA 1997, pp. 418–425 (1997)
13.
Zurück zum Zitat Lanzi, P.L.: An analysis of generalization in the XCS classifier system. Evol. Comput. 7(2), 125–149 (1999)CrossRef Lanzi, P.L.: An analysis of generalization in the XCS classifier system. Evol. Comput. 7(2), 125–149 (1999)CrossRef
14.
Zurück zum Zitat Lanzi, P.L., Loiacono, D., Wilson, S.W., Goldberg, D.E.: XCS with computed prediction in continuous multistep environments. In: IEEE CEC 2005, vol. 3, pp. 2032–2039, September 2005 Lanzi, P.L., Loiacono, D., Wilson, S.W., Goldberg, D.E.: XCS with computed prediction in continuous multistep environments. In: IEEE CEC 2005, vol. 3, pp. 2032–2039, September 2005
15.
Zurück zum Zitat Lanzi, P.L., Loiacono, D., Wilson, S.W., Goldberg, D.E.: Generalization in the XCSF classifier system: analysis, improvement, and extension. Evol. Comput. 15(2), 133–168 (2007)CrossRef Lanzi, P.L., Loiacono, D., Wilson, S.W., Goldberg, D.E.: Generalization in the XCSF classifier system: analysis, improvement, and extension. Evol. Comput. 15(2), 133–168 (2007)CrossRef
16.
Zurück zum Zitat Liu, Y., Browne, W.N., Xue, B.: Absumption to complement subsumption in learning classifier systems. In: Proceedings of GECCO 2019, pp. 410–418. ACM, New York, NY, USA (2019) Liu, Y., Browne, W.N., Xue, B.: Absumption to complement subsumption in learning classifier systems. In: Proceedings of GECCO 2019, pp. 410–418. ACM, New York, NY, USA (2019)
17.
Zurück zum Zitat Liu, Y., Browne, W.N., Xue, B.: Absumption and subsumption based learning classifier systems. In: Proceedings of GECCO 2020, pp. 368–376. ACM, New York, NY, USA (2020) Liu, Y., Browne, W.N., Xue, B.: Absumption and subsumption based learning classifier systems. In: Proceedings of GECCO 2020, pp. 368–376. ACM, New York, NY, USA (2020)
18.
Zurück zum Zitat Pätzel, D., Stein, A., Hähner, J.: A survey of formal theoretical advances regarding XCS. In: Proceedings of GECCO 2019, pp. 1295–1302. Association for Computing Machinery, New York, NY, USA (2019) Pätzel, D., Stein, A., Hähner, J.: A survey of formal theoretical advances regarding XCS. In: Proceedings of GECCO 2019, pp. 1295–1302. Association for Computing Machinery, New York, NY, USA (2019)
19.
Zurück zum Zitat Stein, A., Maier, R., Hähner, J.: Toward curious learning classifier systems: combining XCS with active learning concepts. In: Proceedings of GECCO 2017, pp. 1349–1356. ACM, New York, NY, USA (2017) Stein, A., Maier, R., Hähner, J.: Toward curious learning classifier systems: combining XCS with active learning concepts. In: Proceedings of GECCO 2017, pp. 1349–1356. ACM, New York, NY, USA (2017)
20.
Zurück zum Zitat Stein, A., Maier, R., Rosenbauer, L., Hähner, J.: XCS classifier system with experience replay. In: Proceedings of GECCO 2020, pp. 404–413. ACM, New York, NY, USA (2020) Stein, A., Maier, R., Rosenbauer, L., Hähner, J.: XCS classifier system with experience replay. In: Proceedings of GECCO 2020, pp. 404–413. ACM, New York, NY, USA (2020)
21.
Zurück zum Zitat Stein, A., Menssen, S., Hähner, J.: What about interpolation? A radial basis function approach to classifier prediction modeling in XCSF. In: Proceedings of GECCO 2018, pp. 537–544. ACM, New York, NY, USA (2018) Stein, A., Menssen, S., Hähner, J.: What about interpolation? A radial basis function approach to classifier prediction modeling in XCSF. In: Proceedings of GECCO 2018, pp. 537–544. ACM, New York, NY, USA (2018)
22.
Zurück zum Zitat Stone, C., Bull, L.: For real! XCS with continuous-valued inputs. Evol. Comput. 11(3), 299–336 (2003) Stone, C., Bull, L.: For real! XCS with continuous-valued inputs. Evol. Comput. 11(3), 299–336 (2003)
23.
Zurück zum Zitat Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT press (2018) Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT press (2018)
24.
Zurück zum Zitat Urbanowicz, R.J., Moore, J.H.: Learning classifier systems: a complete introduction, review, and roadmap. J. Artif. Evol. App. 2009, 1:1–1:25 (2009) Urbanowicz, R.J., Moore, J.H.: Learning classifier systems: a complete introduction, review, and roadmap. J. Artif. Evol. App. 2009, 1:1–1:25 (2009)
25.
Zurück zum Zitat Urbanowicz, R.J., Moore, J.H.: ExSTraCS 2.0: description and evaluation of a scalable learning classifier system. Evol. Intell. 8(2), 89–116 (2015) Urbanowicz, R.J., Moore, J.H.: ExSTraCS 2.0: description and evaluation of a scalable learning classifier system. Evol. Intell. 8(2), 89–116 (2015)
26.
Zurück zum Zitat Wilson, S.W.: Generalization in the XCS classifier system. In: Proceedings of GP 1998. Morgan Kaufmann, Madison, Wisconsin, USA (1998) Wilson, S.W.: Generalization in the XCS classifier system. In: Proceedings of GP 1998. Morgan Kaufmann, Madison, Wisconsin, USA (1998)
29.
Zurück zum Zitat Wilson, S.W.: Classifier fitness based on accuracy. Evol. Comput. 3(2), 149–175 (1995)CrossRef Wilson, S.W.: Classifier fitness based on accuracy. Evol. Comput. 3(2), 149–175 (1995)CrossRef
Metadaten
Titel
On the Effects of Absumption for XCS with Continuous-Valued Inputs
verfasst von
Alexander R. M. Wagner
Anthony Stein
Copyright-Jahr
2021
DOI
https://doi.org/10.1007/978-3-030-72699-7_44

Premium Partner