pp 1–26 | Cite as

Feature selection based on community detection in feature correlation networks

  • Miloš SavićEmail author
  • Vladimir Kurbalija
  • Zoran Bosnić
  • Mirjana Ivanović


Feature selection is an important data preprocessing step in data mining and machine learning tasks, especially in the case of high dimensional data. In this paper, we propose a novel feature selection method based on feature correlation networks, i.e. complex weighted networks describing the strongest correlations among features in a dataset. The method utilizes community detection techniques to identify cohesive groups of features in feature correlation networks. A subset of features exhibiting a strong association with the class variable is selected according to the identified community structure taking into account the size of feature communities and connections within them. The proposed method is experimentally evaluated on a high dimensional dataset containing signaling protein features related to the diagnosis of Alzheimer’s disease. We compared the performance of seven commonly used classifiers that were trained without feature selection, after feature selection by four variants of our method determined by different community detection techniques, and after feature selection by four widely used state-of-the-art feature selection methods available in the WEKA machine learning library. The results of the experimental evaluation indicate that our method improves the classification accuracy of several classification models while greatly reducing the dimensionality of the dataset. Additionally, our method tends to outperform traditional feature selection methods provided by the WEKA library.


Feature selection Feature correlation networks Community detection Alzheimer’s disease 



This work is supported by the bilateral project “Intelligent computer techniques for improving medical detection, analysis and explanation of human cognition and behavior disorders” between the Ministry of Education, Science and Technological Development of the Republic of Serbia and the Slovenian Research Agency. M. Savić, V. Kurbalija and M. Ivanović also thank the Ministry of Education, Science and Technological Development of the Republic of Serbia for additional support through Project No. OI174023, “Intelligent techniques and their integration into wide-spectrum decision support”.


  1. 1.
    Blondel VD, Guillaume JL, Lambiotte R, Lefebvre E (2008) Fast unfolding of communities in large networks. J Stat Mech Theory Exp 2008(10):P10008. CrossRefGoogle Scholar
  2. 2.
    Boccaletti S, Latora V, Moreno Y, Chavez M, Hwang DU (2006) Complex networks: structure and dynamics. Phys Rep 424(4–5):175–308. MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Butterworth R, Piatetsky-Shapiro G, Simovici DA (2005) On feature selection through clustering. In: Proceedings of the Fifth IEEE international conference on data mining, ICDM ’05. IEEE Computer Society, Washington, pp. 581–584.
  4. 4.
    Clauset A, Newman MEJ, Moore C (2004) Finding community structure in very large networks. Phys Rev E 70:066111. CrossRefGoogle Scholar
  5. 5.
    Csardi G, Nepusz T (2006) The igraph software package for complex network research. InterJ Complex Syst 1695(5):1–9Google Scholar
  6. 6.
    Duch W (2006) Filter methods. Springer, Berlin, pp 89–117. Google Scholar
  7. 7.
    Fortunato S (2010) Community detection in graphs. Phys Rep 486(35):75–174. MathSciNetCrossRefGoogle Scholar
  8. 8.
    Frank E, Hall M, Holmes G, Kirkby R, Pfahringer B, Witten IH, Trigg L (2010) Weka–a machine learning workbench for data mining. Springer, Boston, pp 1269–1277Google Scholar
  9. 9.
    Guyon I, Elisseeff A (2003) An introduction to variable and feature selection. J Mach Learn Res 3:1157–1182zbMATHGoogle Scholar
  10. 10.
    Hall MA (1998) Correlation-based feature subset selection for machine learning. Ph.D. thesis, University of Waikato, Hamilton, New ZealandGoogle Scholar
  11. 11.
    Horvath S (2011) Correlation and gene co-expression networks. Springer, New York, pp 91–121. Google Scholar
  12. 12.
    Kononenko I (1994) Estimating attributes: analysis and extensions of RELIEF. Springer, Berlin, pp 171–182. Google Scholar
  13. 13.
    Kononenko I, Šimec E, Robnik-Šikonja M (1997) Overcoming the myopia of inductive learning algorithms with RELIEFF. Appl Intell 7(1):39–55. CrossRefGoogle Scholar
  14. 14.
    Krier C, Franois D, Rossi F, Verleysen M (2007) Feature clustering and mutual information for the selection of variables in spectral data. In: Proceedings of European symposium on artificial neural networks advances in computational intelligence and learning, pp 157–162Google Scholar
  15. 15.
    Lal TN, Chapelle O, Weston J, Elisseeff A (2006) Embedded methods. Springer, Berlin, pp 137–165. Google Scholar
  16. 16.
    Li J, Cheng K, Wang S, Morstatter F, Trevino RP, Tang J, Liu H (2016) Feature selection: a data perspective. arXiv preprint arXiv:1601.07996
  17. 17.
    Li Y, Liu W, Jia Y, Dong H (2017) A weighted mutual information biclustering algorithm for gene expression data. Comput Sci Inf Syst 14(3):643–660. CrossRefGoogle Scholar
  18. 18.
    Mann HB, Whitney DR (1947) On a test of whether one of two random variables is stochastically larger than the other. Ann Math Stat 18(1):50–60. MathSciNetCrossRefzbMATHGoogle Scholar
  19. 19.
    Newman MEJ (2003) The structure and function of complex networks. SIAM Rev 45(2):167–256. MathSciNetCrossRefzbMATHGoogle Scholar
  20. 20.
    Newman MEJ (2004) Analysis of weighted networks. Phys Rev E 70:056131. CrossRefGoogle Scholar
  21. 21.
    Newman MEJ, Girvan M (2004) Finding and evaluating community structure in networks. Phys Rev E 69:026113. CrossRefGoogle Scholar
  22. 22.
    Pons P, Latapy M (2006) Computing communities in large networks using random walks. J Graph Algorithms Appl 10(2):191–218. MathSciNetCrossRefzbMATHGoogle Scholar
  23. 23.
    Rand WM (1971) Objective criteria for the evaluation of clustering methods. J Am Stat Assoc 66(336):846–850. CrossRefGoogle Scholar
  24. 24.
    Ray S, Britschgi M, Herbert C, Takeda-Uchimura Y, Boxer A, Blennow K, Friedman L, Galasko D, Jutel M, Karydas A, Kaye J, Leszek J, Miller B, Minthon L, Quinn J, Rabinovici G, Robinson W, Sabbagh M, So Y, Sparks D, Tabaton M, Tinklenberg J, Yesavage J, Tibshirani R, Wyss-Coray T (2007) Classification and prediction of clinical Alzheimer’s diagnosis based on plasma signaling proteins. Nat Med 13(11):1359–1362. CrossRefGoogle Scholar
  25. 25.
    Robnik-Šikonja M, Kononenko I (2003) Theoretical and empirical analysis of ReliefF and RReliefF. Mach Learn 53(1):23–69. CrossRefzbMATHGoogle Scholar
  26. 26.
    Rosvall M, Bergstrom CT (2007) Maps of information flow reveal community structure in complex networks. Proc Natl Acad Sci USA 105(4):1118–1123. CrossRefGoogle Scholar
  27. 27.
    Sánchez-Maroño N, Alonso-Betanzos A, Tombilla-Sanromán M (2007) Filter methods for feature selection—a comparative study. Springer, Berlin, pp 178–187. Google Scholar
  28. 28.
    Savić M, Ivanović M, Radovanović M, Ognjanović Z, Pejović A, Jakšić Krüger T (2015) Exploratory analysis of communities in co-authorship networks: a case study. In: Bogdanova AM, Gjorgjevikj D (eds) ICT innovations 2014. Springer, Cham, pp 55–64. Google Scholar
  29. 29.
    Savić M, Ivanović M, Surla BD (2016) A community detection technique for research collaboration networks based on frequent collaborators cores. In: Proceedings of the 31st annual ACM symposium on applied computing, SAC ’16. ACM, New York, pp 1090–1095.
  30. 30.
    Savić M, Kurbalija V, Ivanović M, Bosnić Z (2017) A feature selection method based on feature correlation networks. In: Ouhammou Y, Ivanovic M, Abelló A, Bellatreche L (eds) Model and data engineering. Springer, Cham, pp 248–261. Google Scholar
  31. 31.
    Slavkov I, Karcheska J, Kocev D, Dzeroski S (2018) HMC-ReliefF: feature ranking for hierarchical multi-label classification. Comput Sci Inf Syst 15(1):187–209. CrossRefGoogle Scholar
  32. 32.
    Song Q, Ni J, Wang G (2013) A fast clustering-based feature subset selection algorithm for high-dimensional data. IEEE Trans Knowl Data Eng 25(1):1–14. CrossRefGoogle Scholar
  33. 33.
    Van Dijck G, Van Hulle MM (2006) speeding up the wrapper feature subset selection in regression by mutual information relevance and redundancy analysis. Springer, Berlin, pp 31–40. Google Scholar
  34. 34.
    Wilcoxon F (1945) Individual comparisons by ranking methods. Biom Bull 1(6):80–83. CrossRefGoogle Scholar
  35. 35.
    Witten IH, Frank E (2005) Data mining: practical machine learning tools and techniques (Morgan Kaufmann Series in Data Management Systems), 2nd edn. Morgan Kaufmann Publishers Inc., San FranciscozbMATHGoogle Scholar
  36. 36.
    Yu L, Liu H (2003) Feature selection for high-dimensional data: A fast correlation-based filter solution. In: Fawcett T, Mishra N (eds) Proceedings of the 20th international conference on machine learning (ICML-03), pp 856–863Google Scholar
  37. 37.
    Zhang Z, Hancock ER (2011) A graph-based approach to feature selection. Springer, Berlin, pp 205–214. zbMATHGoogle Scholar
  38. 38.
    Zhao Z, Liu H (2007) Searching for interacting features. In: Proceedings of the 20th international joint conference on artifical intelligence, IJCAI’07. Morgan Kaufmann Publishers Inc., San Francisco, pp 1156–1161Google Scholar

Copyright information

© Springer-Verlag GmbH Austria, part of Springer Nature 2019

Authors and Affiliations

  • Miloš Savić
    • 1
    Email author
  • Vladimir Kurbalija
    • 1
  • Zoran Bosnić
    • 2
  • Mirjana Ivanović
    • 1
  1. 1.Department of Mathematics and Informatics, Faculty of SciencesUniversity of Novi SadNovi SadSerbia
  2. 2.Faculty of Computer and Information ScienceUniversity of LjubljanaLjubljanaSlovenia

Personalised recommendations