Optimized construction of various classification models for the diagnosis of thyroid problems in human beings
Keywords:
Thyroid disorder, ranked improved F-score ordering, C4.5, MLP, RBFN.Abstract
Thyroid disorder is a major public health problem. Early detection of thyroid disorder is anincreasingly important area in the field of medical diagnosis, pattern recognition, machinelearning and data mining. Thyroid disorder, either over production (hyperthyroidism) or lessproduction (hypothyroidism) results in imbalanced state of thyroid hormone stimulation inhuman beings. So, controlling this disorder has become a central issue in healthcare and needsgreat attention. This research critically examines different classification models constructedusing a novel mathematical ranked improved F-score ordering (RIFO) applied to thyroiddataset taken from machine learning repository, University of California, Irvine. A total ofnine possible and effective feature subsets have been constructed and each subset is testedwith three most benchmarked algorithms namely C4.5, multilayer perceptron (MLP) andradial basis function network (RBFN) using tenfold cross-validation and various training-testpartitions. The obtained results show diverse conclusions, but one with interesting and highestaccuracy has been presented. From the results, it is observed that MLP has emerged with anoutstanding performance of 98.15%, which is greater than all earlier research. The dataset has3 classes, 5 features and 215 records (hypo=30, hyper=35, normal=150).References
Barbaram, B. P., Carrasco-Ochoa, J.A. & Fco, J. 2011. Martinez-Trinidad, General framework for
class-specific feature selection, Expert Systems with Applications 38:10018–10024.
Christopher Jennison. & Bruce W. Turnbull. 2013. Interim monitoring of clinical trials: Decision
theory, dynamic programming and optimal stopping. Kuwait J. Sci. 40(2) pp.43-59.
Coomans, D., Broeckaert, M. Jonckheer M., & Massart D.L. 1983. Comparison of Multivariate
Discriminant Techniques for Clinical Data - Application to the Thyroid Functional State. Meth.
Inform. Med. 22:93-101.
Delen, D., Walker. G. & Kadam. A. 2005. Predicting breast cancer survivability: A Comparison of three
data mining methods. Artificial Intelligence in Medicine. 34:113- 127.
Dogantekin, E., Dogantekin, A. & Avci, D. 2010. ‘An automatic diagnosis system based on thyroid
gland: ADSTG’, Expert Systems with Applications, 37(9) pp.6368–6372.
Dogantekin, E., Dogantekin, A. & Avci, D. 2011. An expert system based on Generalized Discriminant
Analysis and Wavelet Support Vector Machine for diagnosis of thyroid diseases, Expert
Systems with Applications, 38:146-150.
Du, K.-L. & Swamy, M.N.S. 2006. Neural Networks in a Soft computing Framework Springer-
Verlag.
Feyzullah Temurtas. 2009. A comparative study on thyroid disease diagnosis using neural networks’,
Expert Systems with Applications, 36: 944-949.
Fu, X.J. & Wang, L.P. 2003. Data dimensionality reduction with application to simplifying RBF
network structure and improving classification performance, IEEE Transactions on Systems, Man
and Cybernetics-Part B, 33: 399-409.
Hoshi, 2005. An analysis of thyroid function diagnosis using Bayesian-type and SOM- type neural
networks, Chemical and Pharmaceutical Bulletin, 53:1570–1574.
Jaganathan, P. & Rajkumar, N. 2012. An expert system for optimizing thyroid disease diagnosis, Int. J.
Computational Science and Engineering, 7: 232-238
Keles, A. & Keles, A. 2008. ESTDD: Expert system for thyroid diseases Diagnosis, Expert
Systems with Applications. 34: 242–246.
Shu-Hsien, L., Chu, P. & Hsiao, P. 2012. Data mining techniques and applications –A decade review
from 2000 to 2011, Expert Systems with Applications. 39:11303–11311
Mitchell,T. M. 2003. Machine learning, China Machine Press, Beijing.
Nahar, J., Imam, T., Tickle, K.S. & Chen, Y.P. 2013. Computational intelligence for heart disease
diagnosis: A medical knowledge driven approach, Expert Systems with Applications, 40: 96-104
Niloofar, P., Ganjali, M. & Farid Rohani, M.R.2013. Improving the performance of Bayesian networks
in non-ignorable missing data imputation. Kuwait Journal of Science. 40(2) pp.83-101.
Ozyılmaz, L. & Yıldırım, T. 2002. Diagnosis of thyroid disease using artificial neural Network
methods in: Proceedings of ICONIP’02 ninth international conference onneural information
processing, Orchid Country Club, Singapore.
Pasi, L. 2004. Similarity classifier applied to medical data sets, 10 sivua, Fuzziness in Finland’04 in:
International conference on soft computing, Helsinki, Finland & Gulf of Finland & Tallinn,
Estonia.
Paliwal, M. & Kumar, U.A. 2009. Neural networks and statistical techniques: A review of applications,
Expert Systems with Applications, 36:2–17
Polat, K., Sahan, S. & Gunes, S. 2007. A Novel hybrid method based on Artificial Immune
recognition system (AIRS) with fuzzy weighted preprocessing for thyroid Disease diagnosis,
Expert Systems with Applications, 32: 1141–1147.
Quinlan, J.R. 1996. Improved use of continuous attributes in C4.5, Journal of Artificial Intelligence
Research, 4: 77–90.
Serpen, G., Jiang, H. & Allred, L. 1997. Performance analysis of probabilistic potential function
neural network classifier in: Proceedings of artificial neural networks in engineering
conference, St.Louis, MO. 7: 471 – 76.
Xie, J. & Wang, C. 2011. Using support vector machines with a novel feature selection method for
diagnosis of erythemato-squamous diseases, Expert Systems with Applications, 38:5809- 5815.
Yilmaz, I. & Kaynar, O. 2011. Multiple regression, ANN (RBF, MLP) and ANFIS models for prediction
of swell potential of clayey soils, Expert Systems with Applications, 38: 5958–5966.
Zhang, G. & Berardi, L.V. 1998. An investigation of neural networks in thyroid function Diagnosis,
Health Care Management Science. 1:29–37.
