Introduction of Nonlinear Principal Component Analysis with an Application in Health Science Data

Nonlinear Principal Component Analysis is one of the explanatory dimension reducing technique and presents numerical and graphical results for variable set included linear or nonlinear relationships. In this study, Nonlinear Principal Components Analysis was introduced and the relationship between students' sexual and physical trauma stories and demographic characteristics was examined with this method. In the study, the relationship between trauma and 9 variables obtained by questionnaire from 548 students was evaluated by non-linear principal components analysis. The total eigenvalue of first dimension has been found to be 1.766 and the total eigenvalue of second dimension ha s been found to be 1.504 The variance explanation rate of these eigenvalues are 17.656% and 15.044% respectfully. The total explained variance is seen as 28.550%. With nonlinear principal component analysis, categorical variables are scaled to the desired size in the most appropriate way, and thus, nonlinear relationships can be modeled as well as linear relationships between variables. With this analysis, gender, age, marital status and suicide variables were found to be effective on trauma.

PDF

___

1. Timm NH. Applied Multivariate Analysis (2 nd ed). USA: 2002.
2. Leeuw JD. Principal Component Analysis of Binary Data by Iterated Singular Value Decomposition. Computational Statistics and Data Analysis 2006; 50(1): 21-39.
3. Leeuw JD, Rijckevorsel JV. Homals and princals: Some generalizations of principal components analysis. Data Analysis and Informatics 1980; 89: 231-242.
4. Leeuw JD. Nonlinear principal component analysis. Proceedings in Computational Statistics COMPSTAT 1982; October 30 -September 3, Toulouse, 77-86.
5. Hastie T, Stuetzle W. Principal curves. Journal of the American Statistical Association 1989; 84(406): 502-516.
6. Boysan M, Goldsmith RE, Çavuş H, Kayri M, Keskin S. Relations among anxiety, depression and dissociative symptoms, the influence of abuse subtype. Journal of Trauma and Dissociatio 2009; 10: 83-101.
7. Demir Y, Keskin S, Çavuşoğlu Ş. Doğrusal olmayan temel bileşenler analizinin tanıtımı ve uygulanabilirliği. KSÜ Tarım ve Doğa Derg 2021; 24(2): 442-450.
8. SPSS Base 13.0 for Windows. Chicago, IL, SPSS Inc. 2004.
9. Scholz M, Kaplan F, Guy CL, Kopka J, Selbig J. Non-linear PCA: a missing data approach. Bioinformatics 2005; 21(20): 3887-3895.
10. Bekker P, Leeuw JD. Component and Correspondence Analysis (1 nd ed). NewYork: 1988.
11. Burgers D, Leeuw JD. Homogeneity analysis with k sets of variables. An alternating least squares method with optimal scaling features. Psychometrika 1988; 53(2): 177-197.
12. Türe M, Süt N, Yavuz E, Kurt İ, Türe H. Akademik personelin tercihlerin belirlenmesinde, Doğrusal Olmayan Ana Bileşenler Analizi: Tercih haritaları”, Trakya Üniversitesi Bilimsel Araştırmalar Dergisi 2002; 2(1): 72-76.
13. Al-Shaban SA, Al-Faysale MS, Al- Neami AQH. Nonlinear principal component analysis neural network for blind source separation of EEG signals. Research Journal of Applied Sciences, Engineering and Technology 2010; 2(2): 180-190.
14. Gower CJ, Blasius J. Multivariate prediction with nonlinear principal components analysis. Theory, Quality & Quantity 2005; 39: 359-372.
15. DeMers D, Cottrell G. Nonlinear dimensionality reduction. Advances in Neural Information Processing Systems 1993; 5: 580-587.