DOI QR코드

DOI QR Code

Estimating Cumulative Distribution Functions with Maximum Likelihood to Sample Data Sets of a Sea Floater Model

해상 부유체 모델의 표본 데이터에 대해서 최대우도를 갖는 누적분포함수 추정

  • Yim, Jeong-Bin (Division of Maritime Transportation System, College of Maritime Sciences, Mokpo Maritime University) ;
  • Yang, Won-Jae (Division of Maritime Transportation System, College of Maritime Sciences, Mokpo Maritime University)
  • 임정빈 (목포해양대학교 해사대학 해상운송시스템학부) ;
  • 양원재 (목포해양대학교 해사대학 해상운송시스템학부)
  • Received : 2013.07.10
  • Accepted : 2013.09.25
  • Published : 2013.10.31

Abstract

This paper describes evaluation procedures and experimental results for the estimation of Cumulative Distribution Functions (CDF) giving best-fit to the sample data in the Probability based risk Evaluation Techniques (PET) which is to assess the risks of a small-sized sea floater. The CDF in the PET is to provide the reference values of risk acceptance criteria which are to evaluate the risk level of the floater and, it can be estimated from sample data sets of motion response functions such as Roll, Pitch and Heave in the floater model. Using Maximum Likelihood Estimates and with the eight kinds of regulated distribution functions, the evaluation tests for the CDF having maximum likelihood to the sample data are carried out in this work. Throughout goodness-of-fit tests to the distribution functions, it is shown that the Beta distribution is best-fit to the Roll and Pitch sample data with smallest averaged probability errors $\bar{\delta}(0{\leq}\bar{\delta}{\leq}1.0)$ of 0.024 and 0.022, respectively and, Gamma distribution is best-fit to the Heave sample data with smallest $\bar{\delta}$ of 0.027. The proposed method in this paper can be expected to adopt in various application areas estimating best-fit distributions to the sample data.

본 논문에서는 소형 해상 부유체의 위기 평가를 위한 확률기반 위기평가기법(PET)에서 표본 데이터에 최적인 누적분포함수(CDF) 추정에 관한 평가절차와 실험결과를 기술하였다. CDF는 PET에서 부유체의 위기수준을 평가하기 위한 위기허용기준의 참조 값을 제공하기 위한 것으로, 부유체 모델의 롤(Roll), 피치(pitch), 히브(Heave) 등의 운동응답함수에 대한 표본 데이터에서 추정할 수 있다. 본 연구에서는 여덟가지 정형화된 분포함수와 최대우도추정기법을 적용하여 표본 데이터에 대해서 최대우도를 갖는 CDF들을 평가하였다. 분포함수들의 적합도 검정 실험을 통해서, 베타 분포가 롤과 피치 표본 데이터에 대해서 평균 확률오차 $\bar{\delta}(0{\leq}\bar{\delta}{\leq}1.0)$가 가장 작은 0.024와 0.022로 최적임을 나타냈고, 히브 표본 데이터에 대해서는 감마 분포가 $\bar{\delta}$가 가장 작은 0.027로 최적임을 나타냈다. 본 연구에서 제안한 방법은 표본 데이터의 최적분포 추정을 위한 다양한 분야에 적용 가능할 것으로 기대된다.

Keywords

References

  1. Breheny Patrick(2013), Introduction to the empirical distribution function(STA 621): Nonparametric Statistics, white paper, http://web.as.uky.edu/statistics/users/pbreheny/621/F10/notes/8-26.pdf.
  2. Charles R. Farrar, Scott W. Doebling and David A. Nix(2001), "Vibration-based structural damage identification," Philosophical Transactions of the Royal Society A, Londo, Vol. 359, pp. 131-149. https://doi.org/10.1098/rsta.2000.0717
  3. David Vose(2010), Fitting Distributions to Data and why you are probably doing it wrong, white paper, http://www.vosesoftware.com/whitepapers/Fitting%20distributions%20to%20data.pdf.
  4. Joel Azose(2013), On the Gamma Function and Its Applications, white paper, http://www.math.washington.edu/-morrow/336_10/papers/joel.pdf.
  5. Jose Miguel Simon Donaire(2009), Sea Transport Analysis of Upright Wind Turbines, Master Thesis(MEK-FM-EP-2009-14), Technical University of Denmark.
  6. Jun Chang Hyun and Yoo Chul Sang(2012), "Application of the Beta Distribution for the Temporal Quantification of Storm Events," Journal of Korea Water Resources Association, Vol. 45, Issue 6, pp. 531-544. https://doi.org/10.3741/JKWRA.2012.45.6.531
  7. Kim Jin Ho, Kim Hyeong Seok and Cho Sung Ho(2013), "A Ranging Algorithm for IR-UWB in Multi-Path Environment Using Gamma Distribution," The Journal of Korea Information and Communications Society, Vol. 38B, No. 2, pp. 146-153. https://doi.org/10.7840/kics.2013.38B.2.146
  8. Lee J. T. and Oh H. J(1996), "Approximation Equation of Cumulative Distribution Function on the Normal Distribution," Journal of the Korea Society of Mathematical Education, Series A, Vol. 35, No. 1, pp. 57-59, http://www.mathnet.or.kr/mathnet/kms_tex/982256.pdf.
  9. Lu Kung-Chun, Loh Chin-Hsiung, Yang Y. S., Jerome P. Lynch and Kincho H. Law(2008), "Real-Time Structural Damage Detection using Wireless Sensing and Monitoring System," Smart Structures and Systems, TechnoPress, Vol. 4(6), pp. 759-778. https://doi.org/10.12989/sss.2008.4.6.759
  10. MATLAB(2008a), Programming, MATLAB Version 7.6 (R2008a)
  11. MATLAB(2008b), Statistical Toolbox : Maximum likelihood estimation, MATLAB Version 7.6 (R2008a).
  12. MATLAB(2008c), Statistical Toolbox : Empirical Cumulative Distribution Function, MATLAB Version 7.6 (R2008a).
  13. MathWorks(2013), Statistical Toolbox Distribution Functions, http://www.mathworks.co.kr/kr/help/stats/statistics-toolbox-distribution-functions.html.
  14. Plancade Sandra(2013), Adaptive estimation of the conditional cumulative distribution function from current status data, Institute of Community Medicine, white paper, University of Tromso, Norway, pp. 1-42, http://sandraplancade.perso.math.cnrs.fr/cens-int.pdf.
  15. R-forge project(2013), Handbook on probability distributions, white paper, R-forge distributions Core Team, University Year 2009-2010, pp. 1-167, https://r-forge.r-project.org/scm/viewvc.php/*checkout*/pkg/inst/doc/probdistr-main.pdf.
  16. Riddhi D.(2013), Beta Function and its Applications, white paper, Department of Physics and Astronomy, The University of Tennessee, USA, pp. 1-4, http://sces.phys.utk.edu/-moreo/mm08/Riddi.pdf.
  17. Staub Linda and Gekenidis Alexandros(2011), Seminar in Statistics: Survival Analysis Chapter 2, white paper, pp. 1-39, http://stat.ethz.ch/education/semesters/ss2011/seminar/contents/presentation_2.pdf.
  18. Wikipedia(2013a), Tutorial for Probability distribution, http://en.wikipedia.org/wiki/Probability_distribution.
  19. Wikipedia(2013b), Tutorial for Arg max, http://en.wikipedia.org/wiki/Arg_max.
  20. Wikipedia(2013c), Tutorial for Maximum Likelihood, https://en.wikipedia.org/wiki/Maximum_likelihood.
  21. Yim Jeong Bin(2012), "Probability Based Risk Evaluation Techniques for the Small-Sized Sea Floater," Journal of Navigation and Port Research, Vol. 36, No. 10, pp. 795-801. https://doi.org/10.5394/KINPR.2012.36.10.795

Cited by

  1. A Study on the Distribution Estimation of Personal Data Leak Incidents vol.26, pp.3, 2016, https://doi.org/10.13089/JKIISC.2016.26.3.799