The KIPS Transactions:PartB (정보처리학회논문지B)

Korea Information Processing Society (한국정보처리학회)
권호 표지
DOI QR코드

DOI QR Code

Bayesian Network-Based Analysis on Clinical Data of Infertility Patients

베이지안 망에 기초한 불임환자 임상데이터의 분석

  • 정용규 (서울보건대학 전산정보처리과) ;
  • 김인철 (경기대학교 전자계산학과)
  • Published : 2002.10.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we conducted various experiments with Bayesian networks in order to analyze clinical data of infertility patients. With these experiments, we tried to find out inter-dependencies among important factors playing the key role in clinical pregnancy, and to compare 3 different kinds of Bayesian network classifiers (including NBN, BAN, GBN) in terms of classification performance. As a result of experiments, we found the fact that the most important features playing the key role in clinical pregnancy (Clin) are indication (IND), stimulation, age of female partner (FA), number of ova (ICT), and use of Wallace (ETM), and then discovered inter-dependencies among these features. And we made sure that BAN and GBN, which are more general Bayesian network classifiers permitting inter-dependencies among features, show higher performance than NBN. By comparing Bayesian classifiers based on probabilistic representation and reasoning with other classifiers such as decision trees and k-nearest neighbor methods, we found that the former show higher performance than the latter due to inherent characteristics of clinical domain. finally, we suggested a feature reduction method in which all features except only some ones within Markov blanket of the class node are removed, and investigated by experiments whether such feature reduction can increase the performance of Bayesian classifiers.

본 논문에서는 베이지안 망을 기초로 불임환자의 임상 데이터에 대한 다양한 분석 실험을 전개하였다. 이 실험을 통해 임신여부에 영향을 주는 요인들간의 상호의존성을 분석해보고, 또 NBN, BAN, GBN 등 제약조건이 다른 다양한 유형의 베이지안 망 분류기들의 분류성능을 서로 비교해보았다. 그리고 우리는 이와 같은 실험을 통해 임신가능여부(Clin)에 직접적인 영향을 미치는 중요한 요인들로 증상(IND), 약물치료법(stimulation), 여성의 나이(FA), 미세조작 난자의 수(ICT), Wallace 사용여부(ETM) 등 5개의 특성들을 가려낼 수 있었고, 이 요인들간의 상호 의존성도 찾아낼 수 있었다. 또 서로 다른 유형의 베이지안 망 분류기들 중에서 요인들간의 상호의존관계를 허용하는 좀 더 일반적인 BAN과 GBN 등이 그렇지 못한 NBN에 비해 상대적으로 더 높은 분류 성능을 보여준다는 것을 확인하였다. 또 결정트리와 k-최근접 이웃과 같은 다른 분류기들과의 성능 비교를 통해, 임상 데이터의 특성상 확률적 표현과 추론에 기초한 베이지안 망 분류기들이 보다 높은 성능을 보여준다는 사실도 확인할 수 있었다. 또 본 논문에서는 클래스 노드의 Markov blanket에 속한 특성들로 특성집합을 축소하는 것을 제안하고, 실험을 통해 이 특성 축소방법이 베이지안 망 분류기들의 성능을 높여 줄 수 있는지 알아보았다.

Keywords

References

  1. 대한산부인과학회, 부인과학(개정판), 도서출판 칼빈서적, 1991
  2. 정혁, '불임, 무엇이 문제인가 - 그 원인과 치료', 우리출판사, 1997
  3. Bouckaert, R., 'Bayesian Belief Networks : From Construction to Inference,' Doctoral Dissertation, University of Utrecht, The Netherlands, 1995
  4. Cheng, J., Bell, D. A. and Liu, W., 'An Algorithm for Bayesian Belief Network Construction from Data,' Proceedings of AI & STAT-97, Florida, pp.83-90, 1997
  5. Cheng, J. and Greiner, R., 'Learning Bayesian Belief Network Classifiers : Algorithms and System,' Proceedings of the fourteenth Canadian conference on artificial intelligence, 2001
  6. Cheng, J., 'BN PowerConstructor,' http://www.cs. ualberta.ca/~jcheng/bnsoft.htm
  7. Dougherty, J., Kohavi, R., and Sahami, M., 'Supervised and Unsupervised Discretization of Continuous Features,' Proceedings of ICML-95, pp.194-202, 1995
  8. Friedman, N., 'Learning Bayesian Networks in the Presence of Missing Values and Hidden Variables,' Proceedings of ICML-97, pp.125-133, 1997
  9. Friedman, N., Linial, M., Nachman, I., Peter, D., 'Using Bayesian networks to Analyze Expression Data,' Journal of Computational Biology, 2000 https://doi.org/10.1089/106652700750050961
  10. Gorrill, Marsha-J. ; Kaplan, Paul-F. ; Patton, Phillip-E. ; Burry, Kenneth-A., 'Initial Experience with Extended Culture and Blastocyst Transfer of Aryopreserved Embryos,' American Journal of Obstetric & Gynecology, Vol.180, No.6, 1999
  11. Heckerman, D., 'A Tutorial on Learning Bayesian Networks,' Technical Report MSR-TR-95-06, Microsoft Research, 1995
  12. Heckerman, D., Meek, C. and Cooper, G., 'A Bayesian Approach to Causal Discovery,' Technical Report MSR-TR-97-05, Microsoft Research, 1997
  13. Jensen, F. V., An Introduction to Bayesian Networks, New York : Springer-Verlag, 1996
  14. Jiawei Han, Micheline Kamber, Data Mining : Concepts and Techniques, Morgan Kaufmanm. 2001
  15. Kevin Patrick Merphy, 'A Brief Introduction to Graphical Models and Bayesian Networks,' Technical Report, Department of Computer Science, UC Berkley, 2001
  16. Kohavi, R. and John G., 'Wrappers for Feature Subset Selection,' Artificial Intelligence, Special Issue on Relevance, Vol.97, No.1-2, pp.273-324, 1997 https://doi.org/10.1016/S0004-3702(97)00043-X
  17. Langley, P. and Sage, S., 'Induction of Selective Bayesian Classifiers,' Proceedings of UAI-94, 1994
  18. Pazzani, M. J., 'Searching for Dependencies in Bayesian Classifiers,' Proceedings of AI & STAT-95, 1995
  19. Pearl, J., Probabilistic Reasoning in Intelligent Systems, Morgan Kaufmanm, 1988
  20. Provan, G. M. and Singh, M., 'Learning Bayesian Networks Using Feature Selection,' Learning from Data, Lecture Notes in Statistics, Berlin : Springer-Verlag, Vol.112, pp. 291-300, 1996 https://doi.org/10.1007/978-1-4612-2404-4_28
  21. Singh, M., 'Learning Bayesian Networks from Incomplete Data,' Proceedings of AAAI-97, The MIT Press, pp.534-539, 1997
  22. Sprites, P., Gleymour, C, and Sceines, R., Causation, Prediction, and Search, New York : Springer-Verlag, 1993
  23. Tom M. Mitchael, Machine Learning, McGrow-Hill, 1997

Cited by

  1. Bayesian Network-based Data Analysis for Diagnosing Retinal Disease vol.16, pp.3, 2013, https://doi.org/10.9717/kmms.2013.16.3.269