Feature Selection in Taxonomies with Applications to Paleontology

  • Gemma C. Garriga
  • Antti Ukkonen
  • Heikki Mannila
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5255)


Taxonomies for a set of features occur in many real-world domains. An example is provided by paleontology, where the task is to determine the age of a fossil site on the basis of the taxa that have been found in it. As the fossil record is very noisy and there are lots of gaps in it, the challenge is to consider taxa at a suitable level of aggregation: species, genus, family, etc. For example, some species can be very suitable as features for the age prediction task, while for other parts of the taxonomy it would be better to use genus level or even higher levels of the hierarchy. A default choice is to select a fixed level (typically species or genus); this misses the potential gain of choosing the proper level for sets of species separately. Motivated by this application we study the problem of selecting an antichain from a taxonomy that covers all leaves and helps to predict better a specified target variable. Our experiments on paleontological data show that choosing antichains leads to better predictions than fixing specific levels of the taxonomy beforehand.


Feature Selection Leaf Node Target Node Taxonomy Tree Paleontological Data 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Ahuja, R.K., Magnanti, T.L., Orlin, J.B.: Network flows: theory, algorithms, and applications. Prentice-Hall, Inc., Englewood Cliffs (1993)MATHGoogle Scholar
  2. 2.
    Blum, A.L., Langley, P.: Selection of relevant features and examples in machine learning. Artif. Intell. 97(1-2), 245–271 (1997)MathSciNetMATHCrossRefGoogle Scholar
  3. 3.
    Cai, L., Hofmann, T.: Exploiting known taxonomies in learning overlapping concepts. In: IJCAI 2007, pp. 714–719 (2007)Google Scholar
  4. 4.
    Charikar, M., Guruswami, V., Kumar, R., Rajagopalan, S., Sahai, A.: Combinatorial feature selection problems. In: FOCS 2000, page 631 (2000)Google Scholar
  5. 5.
    desJardins, M., Getoor, L., Koller, D.: Using feature hierarchies in Bayesian network learning. In: Choueiry, B.Y., Walsh, T. (eds.) SARA 2000. LNCS (LNAI), vol. 1864, pp. 260–270. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  6. 6.
    Ford, L.R., Fulkerson, D.R.: Maximal flow through a network. Canadian Journal of Mathematics 8, 399–404 (1956)MathSciNetMATHCrossRefGoogle Scholar
  7. 7.
    Fortelius, M., Gionis, A., Jernvall, J., Mannila, H.: Spectral ordering and biochronology of european fossil mammals. Paleobiology 32, 206–214 (2006)CrossRefGoogle Scholar
  8. 8.
    Fortelius, M.: Neogene of the old world database of fossil mammals (NOW) (2008), http://www.helsinki.fi/science/now/
  9. 9.
    Jernvall, J., Fortelius, M.: Common mammals drive the evolutionary increase of hypsodonty in the neogene. Nature 417, 538–540 (2002)CrossRefGoogle Scholar
  10. 10.
    Kohavi, R., John, G.H.: Wrappers for feature subset selection. Artif. Intell. 97(1-2), 273–324 (1997)MATHCrossRefGoogle Scholar
  11. 11.
    Lavrač, N., Gamberger, D.: Relevancy in constraint-based subgroup discovery. In: Constraint-Based Mining and Inductive Databases, pp. 243–266 (2004)Google Scholar
  12. 12.
    Liow, L.H., Fortelius, M., Bingham, E., Lintulaakso, K., Mannila, H., Flynn, L., Stenseth, N.C.: Stenseth higher origination and extinction rates in larger mammals. PNAS 105, 6097–6102 (2008)CrossRefGoogle Scholar
  13. 13.
    Srikant, R., Agrawal, R.: Mining generalized association rules. Future Gener. Comput. Syst. 13(2-3), 161–180 (1997)CrossRefGoogle Scholar
  14. 14.
    Yun, C., Chuang, K., Chen, M.: Using category-based adherence to cluster market-basket data. In: ICDM 2002, p. 546 (2002)Google Scholar
  15. 15.
    Zhang, J., Kang, D.-K., Silvescu, A., Honavar, V.: Learning accurate and concise naïve bayes classifiers from attribute value taxonomies and data. Knowl. Inf. Syst. 9(2), 157–179 (2006)CrossRefGoogle Scholar

Copyright information

© Springer Berlin Heidelberg 2008

Authors and Affiliations

  • Gemma C. Garriga
    • 1
  • Antti Ukkonen
    • 1
  • Heikki Mannila
    • 1
  1. 1.HIITHelsinki University of Technology and University of HelsinkiFinland

Personalised recommendations