İKİ DÜZEYLİ OLASILIK MODELLERİNDE KLASİK VE META SEZGİSEL OPTİMİZASYON TEKNİKLERİNİN PERFORMANSI ÜZERİNE BİR ÇALIŞMA

Bağımlı değişkenin kategorik olduğu durumda, model parametrelerinin tahmininde kullanılan geleneksel yöntem, En Çok Olabilirlik Tahmin Edicisi (EÇOTE)’dir. Bu yöntemde olabilirlik eşitliklerinin çözümünde, klasik Newton-Raphson (NR) algoritması kullanılmaktadır. Ancak bu algoritma olabilirlik fonksiyonunun diferansiyellenebilir özellikte olduğu durum için uygundur. Bu çalışmada, iki düzeyli bağımlı değişken modellerinde klasik optimizasyon yöntemlerinin uygulanabilmesi için gerekli olan varsayımların sağlandığı durumda optimal parametre tahminlerine ulaşabilmek için NR algoritmasına alternatif olan Genetik Algoritma (GA) yaklaşımının etkinliği araştırılmıştır. Bu amaçla, ilk olarak Alopesia hastalığı verisi kullanılmıştır. Gerçek veri uygulamasına ek olarak yapay bir veri kümesi üzerinden elde edilen sonuçlar da sunulmuştur. Son olarak, yöntemlerin Matlab program kodları ve açıklamaları ayrıntılı bir biçimde verilmiştir.

A STUDY ON THE PERFORMANCE OF THE CLASSICAL AND META HEURISTIC OPTIMIZATION TECHNIQUES IN PROBABILITY MODELS WITH TWO LEVELS

The traditional method in the parameter estimation when we study with a categorical dependent variable is the Maximum Likelihood Estimator (MLE). In this method, the classical Newton-Raphson (NR) algorithm is used in the solution of the obtained likelihood equations. However, this algorithm is suitable when the likelihood function is defferentiable. In this study, the efficiency of the Genetic Algorithm approach (GA), alternative to the NR algorithm, is investigated for obtaining the optimal parameter values when the required assumptions for the classical optimization techniques are satisfied in the binary dependent variable models. For this purpose, first, data related to the Alepecia disease is used. In addition to the real data application, the simulated data results are also presented. Finally, the Matlab commands and their explanations are given in detail.

___

Agresti, A., (2002). Categorical Data Analysis. 2th edt., New Jersey, John Wiley&Sons Inc.

Aguilar-Rivera, R., Valenzuela-Rendón, M., Rodríguez-Ortiz, J.J., (2015), “Genetic Algorithms and Darwinian Approaches in Financial Applications: A Survey”, Expert Systems with Applications, 42(21), 7684-7697.

Altunkaynak, B., Esin, A., (2004), “Doğrusal Olmayan Regresyonda Parametre Tahmini İçin Genetik Algoritma Yöntemi”. Gazi Üniversitesi Fen Bilimleri Dergisi, 17(2), 43-51.

Babaoğlu, İ., Findik, O., Ülker, E., (2010), “A Comparison of Feature Selection Models Utilizing Binary Particle Swarm Optimization and Genetic Algorithm in Determining Coronary Artery Disease Using Support Vector Machine”, Expert Systems with Applications, 37(4), 3177-3183.

Goldberg, D.E., (1989). Genetic Algorithms in Search, Optimization and Machine Learning. Addison Wesley, Reading, MA.

Goldberg D.E., Deb, K., (1991), A Comparative Analysis of Selection Schemes Used in Genetic Algorithms, Foundations of Genetic Algorithms., San Francisco, CA: Morgan Kaufmann.

Gordini, N., (2014), “A Genetic Algorithm Approach for Smes Bankruptcy Prediction: Empirical Evidence From Italy”, Expert Systems with Applications, 41(14), 6433-6445.

Hadi, H.S., J.L. Gonzalez-Andujar, (2009), “Comparison of Fitting Weed Seedling Emergence Models With Nonlinear Regression and Genetic Algorithm”, Computers and Electronics in Agriculture, 65(1), 19-25.

Hadji, S., Gaubert, J.P., Krim, F., (2015). “Theoretical and Experimental Analysis of Genetic Algorithms Based MPPT for PV Systems”, Energy Procedia, 74, 772-787.

Holland, J.H., (1975). Adaptation in Natural and Artificial Systems. USA, University of Michigan Press.

Holland, J.H., (1992). Adaptation in Natural and Artificial Systems. 2th edition, Cambridge, London., The MIT Press.

Karr, C.L., Freeman, M. L., (1999). Industrial Applications of Genetic Algorithms., USA, CRC Press.

Koh, Y., Yap, C.W., Li, S.C., (2008). “A Quantitative Approach of Using Genetic Algorithm in Designing A Probability Scoring System of an Adverse Drug Reaction Assessment System”, International Journal of Medical Informatics, 77(6), 421-430.

Johnson, P., Graham, P., Wilson, P., Macaulay, L., Maruff, P., Savage, G., Ellis, K., Martins, R., Rowe, C., Masters, C., Ames, D., Zhang, P., (2013), “Genetic Algorithm with Logistic Regression for Alzheimer's Disease Diagnosis and Prognosis”, Alzheimer's & Dementia, 9(4), P455-P456.

Lee, K.H., Kim, K.W., (2015), “Performance Comparison of Particle Swarm Optimization and Genetic Algorithm for Inverse Surface Radiation Problem”, International Journal of Heat and Mass Transfer, 88, 330-337.

Liu, H.H., Ong, C.S., (2008), “Variable Selection in Clustering for Marketing Segmentation Using Genetic Algorithms”, Expert Systems with Applications, 34(1), 502-510.

Menard, S., (2002). Applied Logistic Regression Analysis, 2th Edition, USA, Sage Publications.

Meng, Q., Weng, J., (2011), “A Genetic Algorithm Approach to Assessing Work Zone Casualty Risk”. Safety Science, 49, 1283-1288.

Mitchell, M., (1999). An Introduction to Genetic Algorithms, 5th Edition, Cambridge, London, The Mit Press.

Pasia, J., Hermosilla, A., Ombao, H., (2005), “A Useful Tool for Statistical Estimation: Genetic Algorithm”, Journal of Statistical Computation and Simulation, 75(4), 237-251.

Pfeifer, J., Barker, K., Ramirez-Marquez, J.E., Morshedlou, N., (2015), “Quantifying the Risk of Project Delays with a Genetic Algorithm”, International Journal of Production Economics, 170(A), 34-44.

Reeves, C.R., Rowe, J.E., (2002). Genetic Algorithms Principles and Perspectives. A Guide to GA Theory., USA., Kluwer Academic Press.

Rechenberg, I., (1973), Evolutions Strategie–Optimierungtechnischersystemenach Prinzipien Der Biologischen Evolution. (PhD.Thesis). Fromman-Holzboog, Germany.

Stylianou, N., Akbarov, A., Kontopantelis, E., Buchan, I., W. Dunn, K., (2015), “Mortality Risk Prediction in Burn Injury: Comparison of Logistic Regression with Machine Learning Approaches”, Burns, 41(5), 925-934.

Yuan, F.C., Lee, C.H., (2015), “Using Least Square Support Vector Regression with Genetic Algorithm to Forecast Beta Systematic Risk”, Journal of Computational Science, 11, 26-33.