Abstract
This paper focuses on the covering mechanism which generates a new if-then rule when the input data does not match the rules in the XCS Classifier System (XCS), a rule-based machine learning system, and discusses how the new rule should be generated from the viewpoint of “inheritance” and “expansion” of the generalization degree of the nearest neighbor rule in the continuous space. For this purpose, this paper proposes the two covering mechanisms based on the “inheritance” and “expansion” of the generalization degree of the nearest neighbor rule and compares their results by applying them to XCS for real-valued input spaces (XCSR). Through the intensive experiments on three types of problems with the different characteristics, the following implications have been revealed: (1) the new rules should be generated by inheriting the generalization degree of the nearest neighbor rule in comparison with expanding it in the continuous space; and (2) XCSR with the “inheritance” based covering mechanism achieves higher classification accuracy with fewer rules than the conventional XCSR, which achieves higher classification accuracy than XCSR with the “expansion” based covering mechanism.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
The experience exp, the time stamp ts, and the action set size as.
- 2.
To implement XCSR-LCGE and XCSR-LCGI, replace the function LocalCovering of XCS-LCPCI in [16] with Algorithm 1.
- 3.
For example, in the case of the bit sequence \(\boldsymbol{b}=11101010101\), \(d=(b_0b_1b_2)_{2}=(111)_{2}=7\), so the correct answer class is determined to be \(b_{k+d}=b_{3+7}=b_{10}=1\).
- 4.
- 5.
11-RMUX: \(N=20,000\), \(\alpha =0.1\), \(\beta =0.229242\), \(\delta =0.1\), \(\nu =5\), \(\theta _{mna}=2\), \(\theta _{GA}=12\), \(\theta _{del}=20\), \(\theta _{sub}=15\), \(\epsilon _0=109.918427\), \(\chi =0.8\), \(\mu =0.04\), \(p_I=0.01\), \(\epsilon _I=0.01\), \(F_I=0.01\), \(FitnessReduction=0.1\), \(m_0=0.1\), \(r_0=1.0\), \(doASSubsumption=yes\), \(doGASubsumption=yes\).
- 6.
6-IRMUX: Analogous to 11-RMUX, except: \(N=2000\), \(\beta =0.030691\), \(\theta _{GA}=192\), \(\theta _{sub}=217\), \(\epsilon _0=9.085638\), \(doASSubsumption=no\).
- 7.
Paddy Leaf: Analogous to 11-RMUX, except: \(N=6400\), \(\beta =0.2\), \(\theta _{mna}=4\), \(\theta _{GA}=48\), \(\theta _{del}=50\), \(\theta _{sub}=50\), \(\epsilon _0=1.0\), \(m_0=0.5\).
- 8.
For example, in the case of the 11-RMUX problem, the generality of each rule in the optimal ruleset [O] [4] is uniformly \(0.5^{k+1}=0.5^4=0.0625\), regardless of the class. Similarly, in the case of 6-IRMUX, it is uniformly \(0.5^{k+1}=0.5^3=0.125\).
References
Barry, A.M.: The stability of long action chains in XCS. Soft. Comput. 6(3–4), 183–199 (2002)
Behdad, M., French, T., Barone, L., Bennamoun, M.: On principal component analysis for high-dimensional XCSR. Evol. Intel. 5(2), 129–138 (2012)
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)
Butz, M.V., Kovacs, T., Lanzi, P.L., Wilson, S.W.: How XCS evolves accurate classifiers. In: Proceedings of the Third Genetic and Evolutionary Computation Conference (GECCO-2001), pp. 927–934. Citeseer (2001)
Butz, M.V., Sastry, K., Goldberg, D.E.: Tournament selection: stable fitness pressure in XCS. In: Cantú-Paz, E., et al. (eds.) GECCO 2003. LNCS, vol. 2724, pp. 1857–1869. Springer, Heidelberg (2003). https://doi.org/10.1007/3-540-45110-2_83
Butz, M.V., Wilson, S.W.: An algorithmic description of XCS. Soft. Comput. 6(3–4), 144–153 (2002)
Fredivianus, N., Prothmann, H., Schmeck, H.: XCS revisited: a novel discovery component for the eXtended classifier system. In: Deb, K., et al. (eds.) SEAL 2010. LNCS, vol. 6457, pp. 289–298. Springer, Heidelberg (2010). https://doi.org/10.1007/978-3-642-17298-4_30
Goldberg, D.E.: Genetic Algorithms in Search, 1st edn. Optimization and Machine Learning. Addison-Wesley Longman Publishing Co., Inc, USA (1989)
Holland, J.H.: Esca** brittleness: The possibilities of general-purpose learning algorithms applied to parallel rule-based systems. Machine learning, an artificial intelligence approach 2, 593–623 (1986)
Kovacs, T.: Towards a theory of strong overgeneral classifiers. In: Foundations of Genetic Algorithms 6, pp. 165–184. Elsevier (2001)
Lanzi, P.L.: An analysis of generalization in the XCS classifier system. Evol. Comput. 7(2), 125–149 (1999)
Nakata, M., Browne, W.N.: Learning optimality theory for accuracy-based learning classifier systems. IEEE Trans. Evol. Comput. 25(1), 61–74 (2020)
Orriols-Puig, A., Bernadó-Mansilla, E.: Bounding XCS’s parameters for unbalanced datasets. In: Proceedings of the 8th Annual Conference on Genetic and Evolutionary Computation, pp. 1561–1568 (2006)
Stone, C., Bull, L.: For real! XCS with continuous-valued inputs. Evol. Comput. 11(3), 299–336 (2003)
Sutton, R.S.: Learning to predict by the methods of temporal differences. Mach. Learn. 3(1), 9–44 (1988)
Tadokoro, M., Hasegawa, S., Tatsumi, T., Sato, H., Takadama, K.: Local covering: adaptive rule generation method using existing rules for XCS. In: 2020 IEEE Congress on Evolutionary Computation (CEC), pp. 1–8. IEEE (2020)
Tadokoro, M., Sato, H., Takadama, K.: XCS with weight-based matching in VAE latent space and additional learning of high-dimensional data. In: 2021 IEEE Congress on Evolutionary Computation (CEC), pp. 304–310. IEEE (2021)
Venturini, G.: Adaptation in dynamic environments through a minimal probability of exploration. In: Proceedings of the Third International Conference on Simulation of Adaptive Behavior: from Animals to Animats 3, pp. 371–379 (1994)
Wada, A., Takadama, K., Shimohara, K., Katai, O.: Analyzing parameter sensitivity and classifier representations for real-valued XCS. In: Kovacs, T., Llorà, X., Takadama, K., Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2003-2005. LNCS (LNAI), vol. 4399, pp. 1–16. Springer, Heidelberg (2007). https://doi.org/10.1007/978-3-540-71231-2_1
Wagner, A.R.M., Stein, A.: On the effects of absumption for XCS with continuous-valued inputs. In: Castillo, P.A., Jiménez Laredo, J.L. (eds.) EvoApplications 2021. LNCS, vol. 12694, pp. 697–713. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-72699-7_44
Widrow, B., Hoff, M.E.: Adaptive switching circuits. Stanford Univ Ca Stanford Electronics Labs, Technical report (1960)
Wilson, S.W.: Classifier fitness based on accuracy. Evol. Comput. 3(2), 149–175 (1995)
Wilson, S.W.: Get real! XCS with continuous-valued inputs. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 1999. LNCS (LNAI), vol. 1813, pp. 209–219. Springer, Heidelberg (2000). https://doi.org/10.1007/3-540-45027-0_11
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 Springer Nature Switzerland AG
About this paper
Cite this paper
Shiraishi, H., Hayamizu, Y., Nakari, I., Sato, H., Takadama, K. (2022). Inheritance vs. Expansion: Generalization Degree of Nearest Neighbor Rule in Continuous Space as Covering Operator of XCS. In: Jiménez Laredo, J.L., Hidalgo, J.I., Babaagba, K.O. (eds) Applications of Evolutionary Computation. EvoApplications 2022. Lecture Notes in Computer Science, vol 13224. Springer, Cham. https://doi.org/10.1007/978-3-031-02462-7_23
Download citation
DOI: https://doi.org/10.1007/978-3-031-02462-7_23
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-02461-0
Online ISBN: 978-3-031-02462-7
eBook Packages: Computer ScienceComputer Science (R0)