Some Paradoxical Perspectives on Approaches and Structures in Functional Data Analysis

Abstract

Developments in functional data analysis have attracted considerable attention in recent literature. This study aims to provide general information about functional data analysis and demonstrate how it can be enriched with auxiliary tools. When the article is considered as a whole, the results predominantly pertain to functional linear models. The first section discusses the estimation of regression model parameters using the Least Squares method. It explains how the Least Squares method is applied within a functional framework and incorporates auxiliary calculations as part of the modeling process. At this stage, the Error Sum of Squares, which forms the basis of the Least Squares method, is represented as a vector field. The second section addresses the interim estimation problem. In this part, the Bernstein polynomial is combined with the wavelet transform to address the interim estimation challenge. The final section introduces various types of functional data analysis. Specifically, the Bernstein polynomial is used in estimating a functional linear model with functional coefficients. Employing the Bernstein polynomial as a model component in the linear model offers a simpler and more innovative approach compared to traditional functional linear model structures. The methods proposed in this study are generally practical and compatible with the classical framework of functional data analysis.

Keywords

• Interpolar estimation
• Kernel estimator
• Longitudinal data
• Bernstein polynomial

References

1. [1] Fisher, R.A., Owen, A.R.G., "An Appreciation of the Life and Work of Sir Ronald Aylmer Fisher: FRS, FSS Sc. D." Journal of the Royal Statistical Society. Series D (The Statistician), 12(4): 313-319, (1962). DOI: https://doi.org/10.2307/2986951
2. [2] Cittert-Eymers, V., "CHD Buys Ballot (1817-1890)." Scientiarum Historia: Tijdschrift voor de Geschiedenis van de Wetenschappen en de Geneeskunde, 10(1): 145-153, (1968).
3. [3] Dozie Kelechukwu, C.N., Christian, C.I., "Decomposition with the Mixed Model in Time Series Analysis using Buys-Ballot Procedure," Asian Journal of Advanced Research and Reports, 17(2): 8-18, (2023). DOI: 10.9734/ajarr/2023/v17i2465
4. [4] Iwueze, I.S., Nwogu. E.C., "Buys–Ballot Estimates for time series decomposition," Global Journal of Mathematical Sciences, 3(2): 83-98, (2004). DOI: 10.4314/gjmas.v3i2.21356
5. [5] Iwueza, I. S., and Ohakwe, J. " Buys-Ballot estimates when stochastic trend is quadratic." Journal of the Nigerian Association of Mathematical Physics, 8: 311-318, (2004). DOI: https://doi.org/10.4314/jonamp.v8i1.40020
6. [6] Doob, Joseph L. "What is a stochastic process?," The American Mathematical Monthly, 49(10): 648-653, (1942). DOI: https://doi.org/10.1080/00029890.1942.11991300
7. [7] Węglarczyk, S., "Kernel density estimation and its application." ITM Web of Conferences. 23. EDP Sciences, (2018). DOI: https://doi.org/10.1051/itmconf/20182300037
8. [8] Leva, J.L., "A fast normal random number generator," ACM Transactions on Mathematical Software (TOMS), 18(4): 449-453, (1992). DOI: https://doi.org/10.1145/138351.138364

9. [9] Joy, K.I., "On-line geometric modeling notes." Computer Science Department, University of California, Davis. http://graphics. cs. UC Davis. Edu/CAGDNotes (1996).
10. [10] Lorentz, G.G., “Bernstein polynomials”, American Mathematical Society, (2012).
11. [11] Gürcan, M., Colak, C., Orman. M.N., “Bernstein polynomial approach against to some frequently used growth curve models on animal data," Pakistan Journal of Statistics, 26(3): 509-516, (2010).
12. [12] Gürcan, M., Colak. C. "Generalization of korovkin type approximation by appropriate random variables & moments and an application in medicine," Pakistan Journal of Statistics, 27(3): 283-297, (2011).
13. [13] Totik, V., "Approximation by Bernstein polynomials," American Journal of Mathematics, 116(4): 995-1018, (1994). DOI: https://doi.org/10.2307/2375007
14. [14] Güral, Y., Demirelli, A., Gürcan, M. "Gölgelerin oyunu: İzdüşümlerin istatistiksel çıkarsamaları ve türkiye’de döviz kurlarını etkileyen makroekonomik göstergeler üzerine bir uygulama," Avrupa Bilim ve Teknoloji Dergisi 38: 341-351, (2022). DOI: https://doi.org/10.31590/ejosat.1039913
15. [15] Aitken, A.C. "IV.—On least squares and linear combination of observations." Proceedings of the Royal Society of Edinburgh, 55: 42-48, (1936). DOI: https://doi.org/10.1017/S0370164600014346
16. [16] Markov, A.A., “Wahrscheinlichkeitsrechnung”. Leipzig. [1287], (1912).
17. [17] Hansen, B.E. "A modern Gauss–Markov theorem," Econometrica, 90(3): 1283-1294, (2022). DOI: https://doi.org/10.3982/ECTA19255
18. [18] Gasser, T., Müller, H.G. "Kernel estimation of regression functions." Smoothing Techniques for Curve Estimation: Proceedings of a Workshop held in Heidelberg, April 2–4, 1979. Springer Berlin Heidelberg, (1979). DOI: https://doi.org/10.1007/BFb0098489
19. [19] Hill. P.D., "Kernel estimation of a distribution function." Communications in Statistics-Theory and Methods, 14(3): 605-620, (1985). DOI: https://doi.org/10.1080/03610928508828937
20. [20] Gürcan, M., Çalık, S., "Generalized convergence characterizations of Feller operator," Pakistan Journal of Statistics, 27(3): 213–219, (2011).
21. [21] Rao, C.R. "Estimation of parameters in a linear model," The Annals of Statistics, 4(6): 1023-1037, (1976). DOI: 10.1214/aos/1176343639
22. [22] Müller, H.G., Stadtmüller, U., "Generalized functional linear models," The Annals of Statistics, 33: 774-805, (2005). DOI: 10.1214/009053604000001156
23. [23] Cardot, H., Ferraty, F., Sarda, P., "Spline estimators for the functional linear model," Statistica Sinica, 13: 571-591, (2003). DOI: https://doi.org/10.1016/S0167-7152(99)00036-X
24. [24] Kwak, N., Choi, C.H. "Input feature selection by mutual information based on Parzen window," IEEE Transactions On Pattern Analysis and Machine Intelligence, 24(12): 1667-1671, (2002). DOI: https://doi.org/10.1109/TPAMI.2002.1114861
25. [25] Babich, G.A., Camps, O.I. "Weighted Parzen windows for pattern classification," IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(5): 567-570, (1996). DOI: https://doi.org/10.1109/34.494647
26. [26] Vincent, P., Bengio. Y., "Manifold parzen windows," Advances in Neural Information Processing Systems, 15: (2002).
27. [27] Köse, H., Gürcan, M., "Zamana bağlı gözlem serilerinin tarihsel başlangıcı ve Buys Ballot’un incelemeleri," Selçuk Üniversitesi Fen Fakültesi Fen Dergisi, 1(18): 33-38, (2001).
28. [28] Ramsay, J.O., Dalzell. C.J., "Some tools for functional data analysis," Journal of the Royal Statistical Society Series B: Statistical Methodology, 53(3): 539-561, (1991). DOI: https://doi.org/10.1111/j.2517-6161.1991.tb01844.x
29. [29] Ullah, S., Caroline F.F. "Applications of functional data analysis: A systematic review," BMC Medical Research Methodology, 13 (2013): 1-12. DOI: https://doi.org/10.1186/1471-2288-13-43
30. [30] Zhu, H., Li, S., Shen, D., Guo, Y., Liu, T., and Chou, Y. C., "FRATS: Functional regression analysis of DTI tract statistics," IEEE Transactions on Medical Imaging, 29(4): 1039-1049, (2010). DOI: https://doi.org/10.1109/TMI.2010.2040625
31. [31] Wu, P.S., Müller, H.G., "Functional embedding for the classification of gene expression profiles," Bioinformatics, 26(4): 509-517, (2010). DOI: https://doi.org/10.1093/bioinformatics/btp711
32. [32] Kim, S.B., Rattakorn, P., Peng, Y.B., "An effective clustering procedure of neuronal response profiles in graded thermal stimulation," Expert Systems with Applications, 37(8): 5818-5826, (2010). DOI: https://doi.org/10.1016/j.eswa.2010.02.025
33. [33] Hyndman, R.J., Shang, H.L., "Rainbow plots, bagplots, and boxplots for functional data," Journal of Computational and Graphical Statistics, 19(1): 29-45, (2010). DOI: https://doi.org/10.1198/jcgs.2009.08158
34. [34] Erbas, B., Sato, T., Yamaguchi, M., Inoue, K., and Fujita, M., "Using functional data analysis models to estimate future time trends in age-specific breast cancer mortality for the United States and England–Wales," Journal of Epidemiology, 20(2): 159-165, (2010). DOI: https://doi.org/10.2188/jea.JE20090072
35. [35] Ghosal, R., Chen, Y., Zhou, Y., and Wang, X., "Shape-constrained estimation in functional regression with Bernstein polynomials," Computational Statistics & Data Analysis, 178: 107614, (2023). DOI: https://doi.org/10.1016/j.csda.2022.107614
36. [36] Bellucci, M. A. "On the explicit representation of orthonormal Bernstein polynomials," arXiv preprint arXiv:1404.2293 (2014). DOI: https://doi.org/10.48550/arXiv.1404.2293