• DocumentCode
    3395209
  • Title

    Research of Refueling customer classifications based on kernel clustering algorithm under the organic combined model

  • Author

    Likun Zhou

  • Author_Institution
    Dept. of Equip. & Transp. Eng., Coll. of CAPF, Xi´an, China
  • fYear
    2011
  • fDate
    19-22 Aug. 2011
  • Firstpage
    1545
  • Lastpage
    1548
  • Abstract
    The kernel clustering algorithm under the organic combined model of the GA and the PSO was put forward. According to the customers´ data of certain, this kernel clustering algorithm was applied to classify the Refueling customers into three groups: the first group includes 3 customers, 5 and 7 customers for the second group and the third group respectively. The result shows that the samples can be classified and the center of mass can be obtained using the data description based on kernel methods. But the clustering region is different when the σ of kernel method gets the different value. This makes the clustering process to be complex, also spends the longer time. While kernel method´s clustering algorithm organically combined with the PSO and GA can get the same result in the shorter time and the calculation is simpler.
  • Keywords
    computer software; customer profiles; customer services; genetic algorithms; particle swarm optimisation; pattern clustering; GA model; PSO model; customer data; kernel clustering algorithm; kernel method based data description; organic combined model; refueling customer classification; Algorithm design and analysis; Classification algorithms; Clustering algorithms; Educational institutions; Genetic algorithms; Kernel; Particle swarm optimization; clustering algorithm; kernel methods; particle swarm algorithm; refueling customer;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Mechatronic Science, Electric Engineering and Computer (MEC), 2011 International Conference on
  • Conference_Location
    Jilin
  • Print_ISBN
    978-1-61284-719-1
  • Type

    conf

  • DOI
    10.1109/MEC.2011.6025768
  • Filename
    6025768