Comparison of Genetic Algorithms and Other Classification Methods in the Diagnosis of Female Urinary Incontinence

Abstract

Galactica, a newly developed machine-learning system that utilizes a genetic algorithm for learning, was compared with discriminant analysis, logistic regression, k-means cluster analysis, a C4.5 decision-tree generator and a random bit climber hill-climbing algorithm. The methods were evaluated in the diagnosis of female urinary incontinence in terms of prediction accuracy of classifiers, on the basis of patient data. The best methods were discriminant analysis, logistic regression, C4.5 and Galactica. Practically no statistically significant differences existed between the prediction accuracy of these classification methods. We consider that machine-learning systems C4.5 and Galactica are preferable for automatic construction of medical decision aids, because they can cope with missing data values directly and can present a classifier in a comprehensible form. Galactica performed nearly as well as C4.5. The results are in agreement with the results of earlier research, indicating that genetic algorithms are a competitive method for constructing classifiers from medical data.

Keywords

Urinary Incontinence - Computer-Assisted Diagnosis - Genetic Algorithms - Machine Learning - Comparison

• References

• 1 Laurikkala J, Juhola M. Learning diagnostic rules from a urological database using a genetic algorithm. In: Alander JT. eds. Proceedings of The Third Nordic Workshop on Genetic Algorithms and their Applications . Helsinki: Finnish Artificial Intelligence Society; 1997: 233-44.
• 2 Laurikkala J, Juhola M. A genetic-based machine learning system to discover the diagnostic rules for female urinary incontinence. Comput Programs Biomed 1998; 55: 217-28.
• 3 Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and Regression Trees . Belmont: Wadsworth International Group; 1984
• 4 Quinlan RJ. C4.5: Programs for Machine Learning . San Mateo: Morgan Kaufmann; 1993
• 5 Armitage P. Statistical Methods in Medical Research . Oxford: Blackwell Scientific Publications; 1994
• 6 Norušis MJ. SPSS for Windows Professional Statistics Release 6.0 . Chicago: SPSS Inc.; 1993
• 7 Whitley D, Rana S, Dzubera J, Mathias KE. Evaluating evolutionary algorithms. Artif Intel 1996; 85: 245-76.
• 8 Davis L. Bit-climbing, representational bias, and test suite design. In: Belew RK, Booker LB. eds. Proceedings of the Fourth International Conference on Genetic Algorithms . San Mateo: Morgan Kaufmann Publishers; 1991: 18-23.
• 9 Wasserman PD. Neural Computing: Theory and Practice . New York: Van Nostrand Reinhold; 1989
• 10 Swingler K. Applying Neural Networks, a Practical Guide . London: Academic Press; 1996
• 11 Holland JH. Adaptation in Natural and Artificial Systems . Ann Arbor: The University of Michigan Press; 1975
• 12 De Jong KA. An Analysis of the Behaviour of a Class of Genetic Adaptive Systems . PhD Thesis. University of Michigan: 1975
• 13 Goldberg DE. Genetic Algorithms in Search, Optimization, and Machine Learning . Reading: Addison-Wesley; 1989
• 14 Mitchell M. An Introduction to Genetic Algorithms . Cambridge, MA: MIT Press; 1996
• 15 Bäck T, Fogel DB, Michalewicz Z. Handbook of Evolutionary Computation . Bristol: Institute of Physics Publishing and Oxford University Press; 1997
• 16 Goldberg DE. Genetic and evolutionary algorithms come of age. Comm ACM 1994; 37 (Suppl. 03) 113-9.
• 17 Fogel DB. The advantages of evolutionary computation. In: Lundh D, Olsson B, Narayanan A. eds. Biocomputing and Emergent Computation: Proceedings of BCEC97 . World Scientific; 1997
• 18 Bäck T, Hammel U, Schwefel H-P. Evolutionary computation: comments on the history and current state. IEEE Transactions on Evolutionary Computation, 1997; 1: 3-17.
• 19 Schwefel H-P. Advantages (and disadvantages) of evolutionary computation over other approaches. In: Bäck T, Fogel BF, and Michalewicz Z. eds. Handbook of Evolutionary Computation . Bristol: Institute of Physics Publishing and Oxford University Press; 1997. A 1.3: 1-2.
• 20 Bonelli P, Parodi A. An efficient classifier system and its experimental comparison with two representative learning methods on three medical domains. In: Belew RK, Booker LB. eds. Proceedings of the Fourth International Conference on Genetic Algorithms . San Mateo: Morgan Kaufmann Publishers; 1991: 288-95.
• 21 Janikow CZ. A knowledge-intensive genetic algorithm for supervised learning. Mach Learn 1993; 13: 189-228.
• 22 Congdon CB. A Comparison of Genetic Algorithms and Other Machine Learning Systems on a Complex Classification Task from Common Disease Research,. PhD Thesis. University of Michigan, Department of Computer Science and Engineering; 1995
• 23 Laurikkala J, Juhola M, Penttinen J, Aukee P. Parameter evaluation of the differential diagnosis of female urinary incontinence for the construction of an expert system. In: Pappas C, Maglaveras N, Scherrer JR. eds. Proceedings of Medical Informatics Europe ‘97 . Amsterdam: IOS Press; 1997: 671-5.
• 24 Laurikkala J, Juhola M, Lammi S, Penttinen J, Aukee P. Statistical evaluation of the parameters in the differential diagnosis of female urinary incontinence for the construction of an expert system. (in press).
• 25 Little RJA, Rubin DB. Statistical Analysis with Missing Data . New York: John Wiley & Sons; 1987
• 26 Schafer JL. Analysis of Incomplete Multivariate Data . London: Chapman & Hall; 1997
• 27 De Jong KA, Spears WA, Gordon DF. Using genetic algorithms for concept learning. Mach Learn 1993; 13: 161-88.
• 28 Michalski RS, Kodratoff Y. Research in machine learning. In: Michalski RS, Kodratoff Y. eds. Machine learning: An Artificial Intelligence Approach, vol. 3 . San Mateo: Morgan Kaufmann Publishers; 1990: 3-30.
• 29 Hart A, Wyatt J. Evaluating black-boxes as medical decision aids: issues arising from a study of neural networks. Med Inf 1990; 15: 229-36.