Evolution of Artificial Neural Network Architecture: Prediction of Depression after Mania

Abstract

Artificial neural networks (ANNs) are compared to standard statistical methods for outcome prediction in biomedical problems. A general method for using genetic algorithms to "evolve" ANN architecture (EANN) is presented. Accuracy of logistic regression, a fully interconnected ANN, and an EANN for predicting depression after mania are examined. All methods showed very good agreement (training set accuracy, chi-square all p <0.01). However, significant differences were found for stability (test set accuracy); logistic regression being the most unstable and EANN being significantly more stable than a fully interconnected ANN (McNemar p <0.01). We conclude that the EANN method enhances ANN stability. This approach may have particular relevance for biomedical prediction problems, such as predicting depression after mania.

Keywords

Artificial Neural Networks - Genetic Algorithms

• REFERENCES

• 1 Baxt WG. Application of neural networks to clinical medicine. Lancet 1995; 346: 1135-8.
• 2 Lucas CP, Rigby JC, Lucas SB. The occurrence of depression following mania. A method of predicting vulnerable cases. J. Psychol 1989; 154: 705-8.
• 3 Minsky M, Papert S. Perceptrons. MIT Press; 1989. Cambridge, Mass, USA.:
• 4 White H. Learning in artificial neural networks: a statistical approach. Neural Comp 1989; 1: 425-64.
• 5 Cross SS, Harrison RF, Kennedy RL. Introduction to neural networks. Lancet 1995; 346: 1075-9.
• 6 Wyatt J. Nervous about artificial neural networks?. Lancet 1995; 346: 1175-7.
• 7 Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Numerical recipes in FORTRAN. The art of scientific computing. 2nd Ed. Cambridge University Press; 1994. Cambridge, UK:
• 8 Hart A, Wyatt J. Evaluating black boxes as medical decision-aids: issues arising from a study of neural networks. Med Inf (Lond) 1990; 15: 229-36.
• 9 Wyatt J, Altman DG. Commentary: Prognostic models: clinically useful or quickly forgotten?. Br Med J 1995; 311: 1539-41.
• 10 Goldberg DE. Genetic algorithms in search, optimization and machine learning. 1st ed. Reading Mass: Addison-Welsey Publ Comp; 1989
• 11 Yao XA. A review of evolutionary artificial neural networks. Int J Intelligent Sys 1993; 8 (4) 539-67.
• 12 Balakrishnan K, Vasant H. Evolutionary design of neural architectures – A preliminary taxonomy and guide to literature. Technical Report CS-TR # 95-01. Iowa State University; Ames, Iowa, USA: 1995
• 13 Branke J. Evolutionary algorithms for neural network design and training. Technical report D-76128. Karlsruhe University; Karlsruhe, Germany: 1995
• 14 Narayanan MN, Lucas SB. A genetic algorithm to improve a neural network to predict a patient's response to warfarin. Meth Inform Med 1993; 32: 55-8.
• 15 Jefferson MF, Pendleton N, Lucas SB, Horan MA. Comparison of a genetic algorithm neural network (GANN) with logistic regression for predicting outcome after surgery for non-small cell lung cancer (NSCLC). Cancer. (in press) 1997
• 16 Dybowski R, Weller P, Chang R, Gant V. Prediction of outcome in critically ill patients using artificial neural network synthesized by genetic algorithm. Lancet 1996; 347: 1146-50.
• 17 Rummelhart DE, Hinton GE, Williams RJ. Learning internal representations by error propagation. In: Rummelhart DE. McClelland (eds).
• 18 Morgan HG. The incidence of depressive symptoms during recovery from hypomania. Br J Psychol 1972; 120: 537-9.
• 19 Cox DR. Analysis of binary data. Methuen Publishers, London, UK, 1970. Parallel distributed processing: explorations in the microstructure of cognition. Vol. 1. MIT Press; Cambridge, MA: 1986
• 20 Swets JA. Measuring the accuracy of diagnostic systems. Science 1988; 240: 1285-93.
• 21 Lawrence S, Giles CL, Tsoi AC. What size neural network gives optimal generalization? Convergence properties of back-propagation. Technical report UMIACS-TR-96-22 and CS-TR-3617, Institute for Advanced Computer Studies. University of Maryland; USA: 1996
• 22 Amari S. Learning and statistical inference, in M. A. Arib, ed., Handbook of Brain Theory and Neural Networks. MIT press; Cambridge, Massachusetts: 1995
• 23 Brown GDA, Hulme C, Hyland PD, Mitchell IJ. Cell suicide in the developing nervous system: a functional neural network model. Cog Brain Res 1994; 2: 71-5.
• 24 Kim J, Ahn J, Cho S. Ensemble competitive learning neural networks with reduced input dimension. Int J Neural Systems 1995; 6: 133-42.
• 25 Dubrovsky B. Fundamental neuroscience and the classification of psychiatric disorders. Neurosci Biobehav Rev 1995; 19: 511-8.
• 26 Lowell WE, Davis GE. Predicting length of stay for psychiatric diagnosis-related groups using neural networks. J Am Med Inform Assoc 1994; 1: 459-66.
• 27 Zou Y, Shen U, Shu L, Wang Y, Feng F, Xu K, Qu Y, Song Y, Zhong Y, Wang M, Lui W. Artificial neural network to assist psychiatric diagnosis. Br J Psychol 1996; 169: 64-7.
• 28 Dumitra A, Radulescu E, Lazarescu V. Improved classification of psychiatric mood disorders using a feedforward neural network. Medinfo 1995; 8: 818-22.
• 29 Jefferson MF, Pendleton N, Lucas SB. Neural Networks. Lancet 1995; 346: 1712.