Araştırma Makalesi
BibTex RIS Kaynak Göster

Yapay Sinir Ağları ile Lojistik Regresyon Analizinin Karşılaştırılması

Yıl 2005, Cilt: 4 Sayı: 1, 57 - 74, 15.04.2005

Öz

Bu çalışmada, öğrencilerin alkol kullanımını etkileyen faktörlerin Lojistik Regresyon Analizi ve Yapay Sinir Ağları ile incelenmesi ve bu yöntemlerin alkol kullanan ve kullanmayan öğrencileri ayırmadaki performanslarının ROC eğrisi yöntemiyle karşılaştırılması amaçlandı.

Çalışmada, 2003-2004 Eğitim-Öğretim yılında Trakya Üniversitesi Tıp Fakültesi 1, 2, 3 ve 4'üncü sınıflarında okuyan öğrencilere Frontal Lob Kişilik Ölçeği ve alkol kullanma alışkanlıklarını tespit etmek için alkolle ilgili anket uygulandı.

Çalışmamızda, Lojistik Regresyon ve dört farklı yapay sinir ağı modeli oluşturuldu. Lojistik Regresyon Analizi sonucunda ders dışındaki zamanlarda bar, disko, kafe ya da kahvehaneye gitme (OR=1.920: p<0.05), dinin önem düzeyi (OR=0.454: p<0.001), alkol kullanan arkadaş sayısı (OR=2.441; p<0.001), alkol içmesi için arkadaşların ısrar düzeyi (OR=1.557; p<0.001) ve dürtüsellik (OR=1.826; p<0.001) değişkenlerinin öğrencilerin alkol kullanımı üzerinde önemli etkiye sahip oldukları bulundu. Lojistik Regresyon Analizi ile Yapay Sinir Ağları ve Yapay Sinir Ağları kendi aralarında karşılaştırıldığında: hiperbolik tanjant-hiperbolik tanjant fonksiyonlu ve hiperbolik-tanjant lojistik fonksiyonlu Yapay Sinir Ağları’nın ROC eğrisi altında kalan ağlarının farklı olmadığı fakat bu modellerin diğer modellerin alanlarından istatistiksel olarak daha büyük oldukları bulundu.

Sonuç olarak; çalışmalarda Yapay Sinir Ağlanrı’nın Lojistik Regresyon Analizi 'ne göre avantaj ve dezavantajları göz önünde bulundurularak amaca göre sınıflandırma ve modelleme çalışmalarının yürütülmesi gerektiğine ve Lojistik Regresyon Analizi 'nin önemsiz değişkenlerin elenmesi için Yapay Sinir Ağları’nda bir eleme yöntemi olarak kullanılabileceğine karar verildi.

Kaynakça

  • AKVARDAR, Y. (2003), Alkol Bağımlılığında Kişilik Özellikleri, Bağımlılık Dergisi, 4(1), 26-30.
  • ARMITAGE, P., BERRY, G. (1994), Statistical Methods in Medical Research, USA: Blackwell Science Ltd.
  • AYCICEGI, A., DINN, W.M., HARRIS, C.L. (2003), Prefrontal Lob Nöropsikolojik Test Bataryası: Sağlıklı Yetişkinlerden Elde Edilen Test Sonuçları, Psikoloji Çalışmaları, 23, 1-26.
  • CHO, H. (2003), Neural Network, Erişim: http://msi.postech.ac.kr/course/ie723/neural-1.pdf Erişim Tarihi: 20.03.2003
  • DİRİCAN, A. (2001), Evaluation of the Diagnostic Test's Performance and Their Comparisons, Cerrahpaşa J Med, 32, 25-30.
  • DREISEITL, S., OHNO-MACHADO, L. (2002), Logistic Regression and Artificial Neural Network Classification Models: A Methodology Review, Journal of Biomedical Infomatics, 35, 352-359.
  • DINN, W.M., AYCICEGI, A., HARRIS, C.L. (2004), Cigarette Smoking in a Student Sample: Neurocognitive and Clinical Correlates, Addictive Behaviors, 29, 107-126.
  • DROBES, J. (2002), Concurrent Alcohol and Tobacco Dependence Mechanisms and Treatment, Alcohol Research & Health, 26(2), 136 -142.
  • EFE, Ö., KAYNAK, O. (2000), Yapay Sinir Ağları ve Uygulamaları, İstanbul: Boğaziçi Üniversitesi.
  • ERGÜN, U., SERHATLIOĞLU, S., HARDALAÇ, F., GÜLER, İ. (2004), Classification of Carotid Artery Stenosis of Patients with Diabetes by Neural Network and Logistic Regression, Computers in Biology and Medicine (In Press).
  • FRANCIS, L. (2001), The Basics of Neural Networks Demystified, Contingencies November/December, 56-61.
  • GAUDART, J., GIUSIANO, B., HUIART, L. (2004), Comparison of the Performance of Multi-Layer Perception and Linear Regression for Epidemiological Data, Computational Statistics&Data Analysis, 44, 547-570.
  • HAJMEER, M., BASHEER, I (2003), Comparison of Logistic Regression and Neural Network Based Classifiers for Bacterial Growth, Food Microbiology, 20, 43-55.
  • HANLEY, J.A., MCNEIL, R.J. (1982), The Meaning and Use of the Area Under a Receiver Operating Characteristic (ROC) Curve, Radiology, 143(1), 29-36.
  • HANLEY, J.A., MCNEIL, B.J. (1983), A Method of Comparing the Areas Under Receiver Operating Characteristic Curves Derived From the Same Cases, Radiology, 148(3), 839-843
  • HASSOUN, M.H. (c1995), Fundamentals of Artificial Neural Networks, Cambridge, Mass: MIT Press.
  • HAYKIN, S. (1999). Neural Network: A Comprehensive Foundation, Upper Saddle River, NJ: Prentice Hall.
  • HEŞEMİNİA, T., ÇALIŞKAN, D., IŞIK , A. (2002), Ankara'da Yüksek Öğrenim Öğrenci Yurtlarında Kalan Öğrencilerin Beslenme Sorunları, İbni Sina Tıp Dergisi, 7, 155-166.
  • HOSMER, D.W., LEMESHOW, S. (2000), Applied Logistic Regression, New York: John Wiley&Sons.
  • KIRKCALDY, B.D., SlEFEN, G., SURALL, D., BISCHOFF, R.J. (2004), Predictors of Drug and Alcohol Abuse Among Children and Adolescents, Personality and Individual Differences, 36, 247-265.
  • KLEINBAUM, D.G. (1994), Logistic Regression: A Self-Learning Text, New York: Springer-Verlag.
  • KÖKNEL, Ö. (2001), Alkol ve Madde Bağımlılığı Alt Kültürü, Bağımlılık Dergisi, 2(2) [http://www.bagimlilik.net/sayi4/alkol¬altkulturu.pdf].
  • KROSE, B., SMAGT, P. (1996), An Introduction to Neural Networks, Amsterdam: The University of Amsterdam.
  • KURT, İ., TÜRE, M., SÜT, N., YAVUZ, E. (2002), Sinir Ağları ile Sistolik Kan Basıncı Değerlerinin Tahmini. Diyarbakır: 6. Biyoistatistik Kongresi.
  • MANEL, S., DIAS, J.M., ORMEROD, S.J. (1999), Comparing Discriminant Analysis. Neural Networks and Logistic Regression for Predicting Species Distiributions: A Case Study with a Himalayan River Bird. Ecological Modelling, 120, 337-347.
  • MCCLISH, D.K. (1987), Comparing the Areas Under More Than Two Independent ROC Curves, Med Decis Making, 7, 149-155.
  • NASR, G.E., BADR, E.A., JOUN, C. (2003), Backpropagation Neural Networks for Modeling Gasoline Consumption , Energy Conversion & Management, 44, 893-905.
  • NGUYEN, T., MALLEY, R., INKELIS, S.H., KUPPERMANN, N. (2002), Comparison of Prediction Models for Adverse Outcome in Pediatric Meningococcal Diseases Using Artificial Neural Network and Logistic Regression Analyses, Journal of Clinical Epidemiology, 55, 687-695.
  • OTTENBACHER, KJ., SMITH, P.M., ILLlG, S.B., LINN, R.T., FlELDER, R.C., GRANGER, C.V. (2001), Comparison of Logistic Regression and Neural Networks to Predict Rehospitalization in Patients with Sroke, Journal of Clinical Epidemiology, 54, 1159-1165.
  • OTTENBACHER, K.J., LINN, R.T., SMITH, P.M., ILLlG, S.B., MANCUSO, M., GRANBER, C.V. (2004), Comparison of Logistic Regression and Neural Network Analysis Applied to Predicting Living Setting After Hip Fracture, Ann Epidemiol (In Press).
  • ÖNCÜ, F., ÖGEL, K., ÇAKMAK, D. (2001), Alkol Kültürü-1: Tarihsel Süreç ve Meyhane Kültürü, Bağımlılık Dergisi, 2(3) [http://www.bagimlilik.net/sayi5/alkol_kulturu.pdf]
  • ÖZDAMAR, K. (1999), Paket Programlarla İstatistiksel Veri Analizi-1, Eskişehir: Kaan Kitabevi.
  • ÖZDAMAR, K. (2003), SPSS ile Biyoistatistik, Eskişehir: Kaan Kitabevi.
  • REMZİ, M., ANAGNOSTOU, T, RAVERY, V., ZLOTTA, A., STEPHAN, C., MARBERGER, M., et al. (2003), An Artificial Neural Networks to Predict the Outcome of Repeat Prostate Biopsies, Adult Urology, 62, 456-460.
  • ROJAS, R. (1991), Neural Networks: A Systematic İntroduction, Berlin: Springer.
  • ROWLAND, T., OHNO-MACHADO, L. , OHRN, A. (1998), Comparison of Multiple Prediction Models for Ambulation Folloving Spinal Cord Injury, Proc. AMIA Annual Symposium: 1998 November 7-11; Orlando, FL, USA, 528-532.
  • SHARMA, S. (1996), Applied Multivariate Techniques, New York: John Wiley&Sons.
  • SUNDARARAJAN, N., SARATCHANDRAN, P. (1998), Parallel Architectures for Artificial Neural Networks: Paradigms and Implementations, California: IEEE Computer Society Press.
  • SWAVING, M., VAN HOUWELINGEN, H., OTTES, F.P., STEERNEMAN, T. (1996), Statistical Comparison of ROC Curves From Multiple Readers, Med Decis Making, 16, 143-152.
  • TATLIDiL, H. (1996), Uygulamalı Çok Değişkenli İstatistiksel Analiz, Ankara: Akademi Matbaası.
  • TU, JV. (1996), Advantages and Disadvantages of Using Artificial Neural Networks Versus Logistic Regression for Predicting Medical Outcomes, J. Clin Epidemiol, 49(11), 1225-1231.
  • TÜRE, M., KURT, İ., YAVUZ, E. (2003), Comparison of Multiple Prediction Models for Degree of Arter Stenosis Determined Angiographically, in: Fredman L, Burgut R, Dafni U (Eds.) EMR 2003. The Second International Biometric Society Conference of the Eastern. Mediterranean Region: 2003 January 12-15; Antalya, Türkiye.
  • TÜRKCAN, A. (2002), Alkol Kullanma İsteğinin (Craving) Mekanizması, Bağımlılık Dergisi, 3(1), 37-42.
  • YAMAMURA, S., KAWADA, K. , TAKEHIRA, R., NISHIZAWA, K., KATAYAMA, S., HIRANO, M. et al. (2004), Artificial Neural Network Modeling to Predict the Plasma Concentration of Aminoglycosides in Burn Patients, Biomedicine & Pharmacotherapy (İn Press).

Comparison of Artificial Neural Networks and Logistic Regression Analysis

Yıl 2005, Cilt: 4 Sayı: 1, 57 - 74, 15.04.2005

Öz

In this study, the factors that affect students' alcohol use behaviors were examined by Logistic Regression Analysis and Artificial Neural Networks. In order to evaluate their success on separation of alcohol user and non-user students, these methods' performance were compared using ROC curve method.

In our study, in order to determine severity of alcohol use among 1, 2, 3 and 4th year students in Trakya University Medical Faculty. 2003-2004, a questionnaire concerning alcohol to predict alcohol use behaviors and Frontal Lobe Personality Scale were performed.

Logistic Regression and four different Artificial Neural Networks models were. Logistic Regression Analysis showed that the following variables effect alcohol use behaviors of students considerably high: to go to bar, disco or cafe in their spare time (OR=1.920, p<0.05), importance level of religion (OR=0.454; p<0.001), the number of alcohol user friends (OR=2.441. p<0.001), the insistence of friends on drinking alcohol (OR=1.557; p<0.01) and impulsivite (OR=1.826, p<0.001). When
Logistic Regression Analysis with Artificial Neural Networks and Artificial Neural Networks each other were compared, there is no difference were observed that the area under the ROC curves of hyperbolic tangent-hyperbolie tangent function and hyperbolic tangent-logistic function Artificial Neural Networks but these models have statistically larger areas than the other models. We could summarize the results of this study as follows: researchers are necessary to take into account advantages and disadvantages of Artificial Neural Networks and Logistic Regression in the classification and modeling, and Logistie Regression Analysis may also use as an elimination methods (in order to eliminate insignificant variables of Logistic Regression) Analysis in Artificial Neural Networks.

Kaynakça

  • AKVARDAR, Y. (2003), Alkol Bağımlılığında Kişilik Özellikleri, Bağımlılık Dergisi, 4(1), 26-30.
  • ARMITAGE, P., BERRY, G. (1994), Statistical Methods in Medical Research, USA: Blackwell Science Ltd.
  • AYCICEGI, A., DINN, W.M., HARRIS, C.L. (2003), Prefrontal Lob Nöropsikolojik Test Bataryası: Sağlıklı Yetişkinlerden Elde Edilen Test Sonuçları, Psikoloji Çalışmaları, 23, 1-26.
  • CHO, H. (2003), Neural Network, Erişim: http://msi.postech.ac.kr/course/ie723/neural-1.pdf Erişim Tarihi: 20.03.2003
  • DİRİCAN, A. (2001), Evaluation of the Diagnostic Test's Performance and Their Comparisons, Cerrahpaşa J Med, 32, 25-30.
  • DREISEITL, S., OHNO-MACHADO, L. (2002), Logistic Regression and Artificial Neural Network Classification Models: A Methodology Review, Journal of Biomedical Infomatics, 35, 352-359.
  • DINN, W.M., AYCICEGI, A., HARRIS, C.L. (2004), Cigarette Smoking in a Student Sample: Neurocognitive and Clinical Correlates, Addictive Behaviors, 29, 107-126.
  • DROBES, J. (2002), Concurrent Alcohol and Tobacco Dependence Mechanisms and Treatment, Alcohol Research & Health, 26(2), 136 -142.
  • EFE, Ö., KAYNAK, O. (2000), Yapay Sinir Ağları ve Uygulamaları, İstanbul: Boğaziçi Üniversitesi.
  • ERGÜN, U., SERHATLIOĞLU, S., HARDALAÇ, F., GÜLER, İ. (2004), Classification of Carotid Artery Stenosis of Patients with Diabetes by Neural Network and Logistic Regression, Computers in Biology and Medicine (In Press).
  • FRANCIS, L. (2001), The Basics of Neural Networks Demystified, Contingencies November/December, 56-61.
  • GAUDART, J., GIUSIANO, B., HUIART, L. (2004), Comparison of the Performance of Multi-Layer Perception and Linear Regression for Epidemiological Data, Computational Statistics&Data Analysis, 44, 547-570.
  • HAJMEER, M., BASHEER, I (2003), Comparison of Logistic Regression and Neural Network Based Classifiers for Bacterial Growth, Food Microbiology, 20, 43-55.
  • HANLEY, J.A., MCNEIL, R.J. (1982), The Meaning and Use of the Area Under a Receiver Operating Characteristic (ROC) Curve, Radiology, 143(1), 29-36.
  • HANLEY, J.A., MCNEIL, B.J. (1983), A Method of Comparing the Areas Under Receiver Operating Characteristic Curves Derived From the Same Cases, Radiology, 148(3), 839-843
  • HASSOUN, M.H. (c1995), Fundamentals of Artificial Neural Networks, Cambridge, Mass: MIT Press.
  • HAYKIN, S. (1999). Neural Network: A Comprehensive Foundation, Upper Saddle River, NJ: Prentice Hall.
  • HEŞEMİNİA, T., ÇALIŞKAN, D., IŞIK , A. (2002), Ankara'da Yüksek Öğrenim Öğrenci Yurtlarında Kalan Öğrencilerin Beslenme Sorunları, İbni Sina Tıp Dergisi, 7, 155-166.
  • HOSMER, D.W., LEMESHOW, S. (2000), Applied Logistic Regression, New York: John Wiley&Sons.
  • KIRKCALDY, B.D., SlEFEN, G., SURALL, D., BISCHOFF, R.J. (2004), Predictors of Drug and Alcohol Abuse Among Children and Adolescents, Personality and Individual Differences, 36, 247-265.
  • KLEINBAUM, D.G. (1994), Logistic Regression: A Self-Learning Text, New York: Springer-Verlag.
  • KÖKNEL, Ö. (2001), Alkol ve Madde Bağımlılığı Alt Kültürü, Bağımlılık Dergisi, 2(2) [http://www.bagimlilik.net/sayi4/alkol¬altkulturu.pdf].
  • KROSE, B., SMAGT, P. (1996), An Introduction to Neural Networks, Amsterdam: The University of Amsterdam.
  • KURT, İ., TÜRE, M., SÜT, N., YAVUZ, E. (2002), Sinir Ağları ile Sistolik Kan Basıncı Değerlerinin Tahmini. Diyarbakır: 6. Biyoistatistik Kongresi.
  • MANEL, S., DIAS, J.M., ORMEROD, S.J. (1999), Comparing Discriminant Analysis. Neural Networks and Logistic Regression for Predicting Species Distiributions: A Case Study with a Himalayan River Bird. Ecological Modelling, 120, 337-347.
  • MCCLISH, D.K. (1987), Comparing the Areas Under More Than Two Independent ROC Curves, Med Decis Making, 7, 149-155.
  • NASR, G.E., BADR, E.A., JOUN, C. (2003), Backpropagation Neural Networks for Modeling Gasoline Consumption , Energy Conversion & Management, 44, 893-905.
  • NGUYEN, T., MALLEY, R., INKELIS, S.H., KUPPERMANN, N. (2002), Comparison of Prediction Models for Adverse Outcome in Pediatric Meningococcal Diseases Using Artificial Neural Network and Logistic Regression Analyses, Journal of Clinical Epidemiology, 55, 687-695.
  • OTTENBACHER, KJ., SMITH, P.M., ILLlG, S.B., LINN, R.T., FlELDER, R.C., GRANGER, C.V. (2001), Comparison of Logistic Regression and Neural Networks to Predict Rehospitalization in Patients with Sroke, Journal of Clinical Epidemiology, 54, 1159-1165.
  • OTTENBACHER, K.J., LINN, R.T., SMITH, P.M., ILLlG, S.B., MANCUSO, M., GRANBER, C.V. (2004), Comparison of Logistic Regression and Neural Network Analysis Applied to Predicting Living Setting After Hip Fracture, Ann Epidemiol (In Press).
  • ÖNCÜ, F., ÖGEL, K., ÇAKMAK, D. (2001), Alkol Kültürü-1: Tarihsel Süreç ve Meyhane Kültürü, Bağımlılık Dergisi, 2(3) [http://www.bagimlilik.net/sayi5/alkol_kulturu.pdf]
  • ÖZDAMAR, K. (1999), Paket Programlarla İstatistiksel Veri Analizi-1, Eskişehir: Kaan Kitabevi.
  • ÖZDAMAR, K. (2003), SPSS ile Biyoistatistik, Eskişehir: Kaan Kitabevi.
  • REMZİ, M., ANAGNOSTOU, T, RAVERY, V., ZLOTTA, A., STEPHAN, C., MARBERGER, M., et al. (2003), An Artificial Neural Networks to Predict the Outcome of Repeat Prostate Biopsies, Adult Urology, 62, 456-460.
  • ROJAS, R. (1991), Neural Networks: A Systematic İntroduction, Berlin: Springer.
  • ROWLAND, T., OHNO-MACHADO, L. , OHRN, A. (1998), Comparison of Multiple Prediction Models for Ambulation Folloving Spinal Cord Injury, Proc. AMIA Annual Symposium: 1998 November 7-11; Orlando, FL, USA, 528-532.
  • SHARMA, S. (1996), Applied Multivariate Techniques, New York: John Wiley&Sons.
  • SUNDARARAJAN, N., SARATCHANDRAN, P. (1998), Parallel Architectures for Artificial Neural Networks: Paradigms and Implementations, California: IEEE Computer Society Press.
  • SWAVING, M., VAN HOUWELINGEN, H., OTTES, F.P., STEERNEMAN, T. (1996), Statistical Comparison of ROC Curves From Multiple Readers, Med Decis Making, 16, 143-152.
  • TATLIDiL, H. (1996), Uygulamalı Çok Değişkenli İstatistiksel Analiz, Ankara: Akademi Matbaası.
  • TU, JV. (1996), Advantages and Disadvantages of Using Artificial Neural Networks Versus Logistic Regression for Predicting Medical Outcomes, J. Clin Epidemiol, 49(11), 1225-1231.
  • TÜRE, M., KURT, İ., YAVUZ, E. (2003), Comparison of Multiple Prediction Models for Degree of Arter Stenosis Determined Angiographically, in: Fredman L, Burgut R, Dafni U (Eds.) EMR 2003. The Second International Biometric Society Conference of the Eastern. Mediterranean Region: 2003 January 12-15; Antalya, Türkiye.
  • TÜRKCAN, A. (2002), Alkol Kullanma İsteğinin (Craving) Mekanizması, Bağımlılık Dergisi, 3(1), 37-42.
  • YAMAMURA, S., KAWADA, K. , TAKEHIRA, R., NISHIZAWA, K., KATAYAMA, S., HIRANO, M. et al. (2004), Artificial Neural Network Modeling to Predict the Plasma Concentration of Aminoglycosides in Burn Patients, Biomedicine & Pharmacotherapy (İn Press).
Toplam 44 adet kaynakça vardır.

Ayrıntılar

Birincil Dil Türkçe
Konular İstatistik
Bölüm Araştırma Makaleleri
Yazarlar

İmran Kurt Omurlu

Mevlüt Türe Bu kişi benim

Yayımlanma Tarihi 15 Nisan 2005
Yayımlandığı Sayı Yıl 2005 Cilt: 4 Sayı: 1

Kaynak Göster

APA Kurt Omurlu, İ., & Türe, M. (2005). Yapay Sinir Ağları ile Lojistik Regresyon Analizinin Karşılaştırılması. İstatistik Araştırma Dergisi, 4(1), 57-74.