Comparative study on numerical methods for maximum likelihood estimation and development of MLE-SYS system 最尤推定法のための数値解法の比較研究とMLE-SYSシステムの研究開発

Access this Article
Search this Article
Author
    • 道家, 暎幸, 1944- ドウケ, ヒデユキ
Bibliographic Information
Title

Comparative study on numerical methods for maximum likelihood estimation and development of MLE-SYS system

Other Title

最尤推定法のための数値解法の比較研究とMLE-SYSシステムの研究開発

Author

道家, 暎幸, 1944-

Author(Another name)

ドウケ, ヒデユキ

University

九州大学

Types of degree

理学博士

Grant ID

乙第4919号

Degree year

1991-03-27

Note and Description

博士論文

Chapter1 PRELIMINARIES Chapter2 UNCONSTRAINED METHODS IN NUMERICAL ANALYSIS Chapter3 CONSTRAINED METHODS IN NUMERICAL ANALYSIS Chapter4 COMPARATIVE STUDY ON NUMERICAL METHODS FOR ESTIMATING PARAMETERS OF NORMAL MIXTURE DISTRIBUTION Chapter5 COMPARATIVE STUDY ON NUMERICAL METHODS FOR ESTIMATING PARAMETERS OF MULTIVARIATE EXPONENTIAL DISTRIBUTION Chapter6 DESIGN AND DEVELOPMENT OF MLE-SYS Chapter7 APPLICATION OF MLE-SYS FOR ESTIMATING PARAMETERS OF LATENT CLASS MODEL Chapter8 FURTHER DISCUSSION

主1-参1

Table of Contents
  1. CONTENTS / (0005.jp2)
  2. PREFACE / (0003.jp2)
  3. Chapter1 PRELIMINARIES / p1 (0008.jp2)
  4. 1.1 Maximum Likelihood Estimation / p2 (0009.jp2)
  5. 1.2 Weibull Distribution / p3 (0010.jp2)
  6. 1.3 Mixture Normal Distribution / p5 (0012.jp2)
  7. 1.4 Multivariate Exponential Distribution / p6 (0013.jp2)
  8. 1.5 Latent Class Model / p7 (0014.jp2)
  9. Chapter2 UNCONSTRAINED METHODS IN NUMERICAL ANALYSIS / p11 (0018.jp2)
  10. 2.1 Introduction / p11 (0018.jp2)
  11. 2.2 Steepest Descent Method / p12 (0019.jp2)
  12. 2.3 Newton-Raphson Method / p14 (0021.jp2)
  13. 2.4 Levenberg-Marquardt Method / p16 (0023.jp2)
  14. 2.5 Variable Metric Methods / p16 (0023.jp2)
  15. 2.6 Conjugate Gradient Methods / p17 (0024.jp2)
  16. 2.7 Powell Methods / p18 (0025.jp2)
  17. 2.8 Direct Search Methods / p19 (0026.jp2)
  18. Chapter3 CONSTRAINED METHODS IN NUMERICAL ANALYSIS / p22 (0029.jp2)
  19. 3.1 Introduction / p22 (0029.jp2)
  20. 3.2 Penalty Method / p23 (0030.jp2)
  21. 3.3 Barrier Method / p23 (0030.jp2)
  22. 3.4 Multiplier Method / p24 (0031.jp2)
  23. 3.5 Gradient Projection Method / p24 (0031.jp2)
  24. 3.6 Sequential Quadratic Programming Method / p24 (0031.jp2)
  25. Chapter4 COMPARATIVE STUDY ON NUMERICAL METHODS FOR ESTIMATING PARAMETERS OF NORMAL MIXTURE DISTRIBUTION / p27 (0034.jp2)
  26. 4.1 ML Estimation of Normal Mixture Distribution / p27 (0034.jp2)
  27. 4.2 Applied Methods in Numerical Analysis / p28 (0035.jp2)
  28. 4.3 Comparative Evaluations / p45 (0052.jp2)
  29. Chapter5 COMPARATIVE STUDY ON NUMERICAL METHODS FOR ESTIMATING PARAMETERS OF MULTIVARIATE EXPONENTIAL DISTRIBUTION / p46 (0053.jp2)
  30. 5.1 ML Estimation of Multivariate exponential distribution / p46 (0053.jp2)
  31. 5.2 Applied Methods in Numerical Analysis / p48 (0055.jp2)
  32. 5.3 Comparative Evaluations / p58 (0065.jp2)
  33. Chapter6 DESIGN AND DEVELOPMENT OF MLE-SYS / p59 (0066.jp2)
  34. 6.1 Features of System / p59 (0066.jp2)
  35. 6.2 Configuration of MLE-SYS / p61 (0068.jp2)
  36. 6.3 System Flow / p65 (0072.jp2)
  37. 6.4 Procedure on Likelihood Estimation / p70 (0077.jp2)
  38. 6.5 Illustrative Operations / p73 (0080.jp2)
  39. Chapter7 APPLICATION OF MLE-SYS FOR ESTIMATING PARAMETERS OF LATENT CLASS MODEL / p77 (0084.jp2)
  40. 7.1 A Latent Class Model and Gibson's Method / p77 (0084.jp2)
  41. 7.2 ML Method and N.R. Method / p80 (0087.jp2)
  42. 7.3 Proposed Procedure on the Basis of MLE-SYS / p83 (0090.jp2)
  43. 7.4 Comparative Evaluations / p85 (0092.jp2)
  44. Chapter8 FURTHER DISCUSSION / p87 (0094.jp2)
  45. ACKNOWLEDGMENTS / p90 (0097.jp2)
  46. APPENDIX / p91 (0098.jp2)
  47. REFERENCES / p101 (0108.jp2)
5access

Codes
  • NII Article ID (NAID)
    500000075616
  • NII Author ID (NRID)
    • 8000000998423
  • DOI(NDL)
  • Text Lang
    • eng
  • NDLBibID
    • 000000239930
  • Source
    • Institutional Repository
    • NDL ONLINE
    • NDL Digital Collections
Export
Page Top