International E-publication: Publish Projects, Dissertation, Theses, Books, Souvenir, Conference Proceeding with ISBN.  International E-Bulletin: Information/News regarding: Academics and Research

A Generalized Multi-Group Discriminant Procedure for Classification: A Comparative Study

    , , , , ,

Author Affiliations

  • 1Department of Mathematics and Statistics, Federal Polytechnic, P. M. B. 55, Bida, Niger State, NIGERIA
  • 2Department of Statistics, Imo State University, P.M.B. 2000, Owerri, Imo State, NIGERIA
  • 3Department of Statistics, Nnamdi Azikiwe University, Awka, Anambra State, NIGERIA
  • 4Department of Mathematics/Computer Science/Statistics/Informatics, Federal University, Ndufu-Alike Ikwo, P.M.B. 1010, Abakaliki, Ebonyi State, NIGERIA

Res. J. Mathematical & Statistical Sci., Volume 3, Issue (12), Pages 1-4, December,12 (2015)

Abstract

It has been earlier demonstrated that an alternative and dependable tool for discriminant analysis with many groups is obtainable by considering all possible pairs of group of the available multi-groups. An assessment of the performance of this procedure is therefore made by comparing its accuracy rate alongside other conventional and common procedures of classification, with a distributional data set under various sample sizes. The All Possible Pairs (APP) classification procedure performed better than its conventional counterparts under the various scenario. Thus, the All Possible Pairs procedure could still remain a better option in situations of any multivariate data structure with many groups.

References

  1. Ekezie D.D., Principal Component Analysis, an Aid toInterpretation of Data: A Case Study of Oil Palm (Elaeisguineensis jacq), J. of Emerging Trends in Engr. andApp. Sci., 4(2), 237-241 (2013)
  2. Onyeagu S.I., Osuji G.A. and Onyia O.M., Comparisonof Canonical Correlation Analysis and the GeneralizedCanonical Correlation Analysis using the Lognormal andCauchy Distributed Data, J. of Math. Theory and Mod.,5(4),177-185 (2014)
  3. Fisher R.A., The Use of Multiple Measurements inTaxonomic Problems, Ann. Eugenics., (7), 179-188(1936)
  4. Fisher R.A., The Precision of Discriminant Functions, Ann. Eugenics., (10), 422-429 (1940)
  5. Fukunaga F., Introduction to statistical patternrecognition, Academic Press, (1990)
  6. Sandrine D., Jane F. and Terence P.S., Comparison ofDiscrimination Methods for the Classification of TumorsUsing Gene Expression Data. Technical Report, 576,Department of Statistics
  7. Breiman L., Bagging Predictors, Machine Learning., (24),123-140 (1996)
  8. Breiman L., Arcing Classifiers, Ann. of Stat., (26), 801-824 (1998)
  9. Freund Y. and Schapire R.E., A Decision-TheoreticGeneralization of On-line Learning and an Applicationto Boosting, J. of Comp. and System Sci., (55), 119-139(1997)
  10. Lesaffre E. and Albert A., Multiple-Group LogisticRegression Diagnostics, J. of the Royal Stat. Soc., 38(3),425-440 (1989)
  11. Vapnik V.N., Statistical Learning Theory, Wiley, NewYork., (1998)
  12. Tao L., Shenghou Z. and Mitsunori O., UsingDiscriminant Analysis for Multi-class Classification: AnExperimental Investigation, Knowledge and InformationSystems., 10(4), 453-472 (2006)
  13. Hastie T. and Tibshirani R., Classification by PairwiseCoupling, In M. I. Jordan, M. J. Kearns, and S. A. Solla(eds.), Advances in Neural Information ProcessingSystems, 10, (1998)
  14. Kwon Y.S., Han I.G. and Lee K.C., Ordinal PairwisePartitioning approach to Neural Networks, Training inBond Rating, Intelligent Systems in Accounting Financeand Management., (6), 23-40 (1997)
  15. Platt J.C., Cristianini N. and Taylor S.J., Large MarginDAGs for Multiclass Classification, In S. Solla., T.Leen, and K.R. Muller (eds.)
  16. Dietterich T.G. and Bakiri G., Solving MulticlassLearning Problems via Error-Correcting Output Codes, J. of Artif. Intell. Res., (2), 263–286 (1995)
  17. Anyanwu P.E., Ekezie D.D. and Onyeagu S.I., A Reviewof the limitations of some Discriminant Analysisprocedures in Multiple Group Classification, J. of MathTheory and Modeling., 5(9),199-203 (2015)
  18. David H., Error-Rate Estimation in Multiple-GroupLinear Discriminant Analysis, Technometrics., 4(38),389-399 (1996)
  19. Anyanwu P.E., Ekezie D.D., Onyeagu. S.I. andNwankwo I.I.M., A Generalized Multi-GroupDiscriminant Function procedure for Classification: AnApplication to Ten Groups of Yam Species, J. of MathTheory and Modeling., 5(10),115-123 (2015)
  20. Oyeyemi G.M. and Mbaeyi G.C., On Discriminationprocedure with mixtures of Continuous and CategoricalVariables, Unpublished M.Sc Thesis, Department ofStatistics, University of Ilorin, Ilorin, Nigeria, (2013)