Communications for Statistical Applications and Methods

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

Graphical Methods for the Sensitivity Analysis in Discriminant Analysis

  • Jang, Dae-Heung (Department of Statistics, Pukyong National University) ;
  • Anderson-Cook, Christine M. (Statistical Sciences Group, Los Alamos National Laboratory) ;
  • Kim, Youngil (School of Business and Economics, Chung-Ang University)
  • Received : 2015.06.28
  • Accepted : 2015.07.28
  • Published : 2015.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

Similar to regression, many measures to detect influential data points in discriminant analysis have been developed. Many follow similar principles as the diagnostic measures used in linear regression in the context of discriminant analysis. Here we focus on the impact on the predicted classification posterior probability when a data point is omitted. The new method is intuitive and easily interpretable compared to existing methods. We also propose a graphical display to show the individual movement of the posterior probability of other data points when a specific data point is omitted. This enables the summaries to capture the overall pattern of the change.

Keywords

References

  1. Aitchison, J. and Dunsmore, I. R. (1975). Statistical Prediction Analysis, Cambridge University Press, Cambridge, UK.
  2. Campbell, M. J. (2001). Statistics at Square Two: Understanding Modern Statistical Applications in Medicine, Boston Medical Publishers Inc., London.
  3. Campbell, N. A. (1978). The influence function as an aid in outlier detection in discriminant analysis, Applied Statistics, 27, 251-258. https://doi.org/10.2307/2347160
  4. Cook, R. D. (1977). Detection of influential observations in linear regression, Technometrics, 19, 15-18.
  5. Cook, R. D. (1986). Assessment of local influence, Journal of the Royal Statistical Society Series B (Methodological), 48, 133-169.
  6. Critchley, F. and Vitiello, C. (1991). The influence of observations on misclassification probability estimates in linear discriminant analysis, Biometrika, 78, 677-690. https://doi.org/10.1093/biomet/78.3.677
  7. Fung, W. K. (1992). Some diagnostic measures in discriminant analysis, Statistics & Probability Letters, 13, 279-285. https://doi.org/10.1016/0167-7152(92)90035-4
  8. Fung, W. K. (1995). Diagnostics in linear discriminant analysis, Journal of the American Statistical Association, 90, 952-956. https://doi.org/10.1080/01621459.1995.10476595
  9. Jung, K. M. (1998). Local influence assessment of the misclassification probability in multiple discriminant analysis, Journal of the Korean Statistical Society, 27, 471-483.
  10. Lachenbruch, P. A. (1997). Discriminant diagnostics, Biometrics, 53, 1284-1292. https://doi.org/10.2307/2533497
  11. Lahiff, M. and Whitcomb, K. M. (1990). Empirical influence function for misclassification rates in discriminant analysis, Communication in Statistics - Theory and Methods, 19, 2999-3009. https://doi.org/10.1080/03610929008830360
  12. Lee, H. J. and Kim, H. G. (2011). Derivation and application of influence function in discriminant analysis for three groups, The Korean Journal of Applied Statistics, 24, 941-949. https://doi.org/10.5351/KJAS.2011.24.5.941
  13. Moreno-Roldan, D., Munoz-Pichardo, J. M. and Enguix-Gonzales, A. (2007). Influence diagnostics in multiple discriminant analysis, Test, 16, 172-187. https://doi.org/10.1007/s11749-006-0007-9
  14. Poon, W. Y. (2004). Identifying influential observations in discriminant analysis, Statistical Methods in Medical Research, 13, 291-308.