해양환경에 관련된 실험을 위한 최적실험계획

Optimal Designs for the Experiments related with Marine Environment

  • 김재환 (한국해양대학교 이공대학 응용수학과)
  • 발행 : 1997.07.01

초록

This paper develops a new heuristic, the Excursion Algorithm(EA), for constructing optimal designs for the experiments related with marine environment. The proposed EA consists of three parts: 1) construction of an initial feasible solution, 2) excursions over a bounded region, and 3) stopping rules. It is the second part that distinguishes the EA from the other existing heuristic methods. It turns out that excursions over a bounded feasible and/or infeasible region is effective in alleviating the risks of being trapped at a local optimum. Since this problem is formulated for the first time thesis, other heuristic algorithms do not exist. Therefore, global optimal solutions are obtained by complete enumeration for some cases, and the performance of the EA is evaluated in terms of solution quality. Computational results show that the proposed EA is effective in finding good(or, in many cases, global) solutions to the constrained optimal experimental design problems.

키워드

참고문헌

  1. Biometrika v.12 On the standard deviation of adjusted and inter-polated values of an observed polynomial function Smith, K.
  2. Annals of Mathematical Statistics v.14 On the efficient design of statistical investigations Wald, A.
  3. Annals of Mathematical Statistics v.15 Some improvements in weighing and other experimental techniques Hotelling, H.
  4. Annals of Mathematical Statistics v.17 On Hotelling's weighing problem Mood, A.M.
  5. Canadian Journal of Mathmatics v.12 The equivalence of two extremum problem Kiefer, J.;Wolfowitz, J.
  6. International Statistical Review v.50 Developments in the design of experiments Atkinson, A.C.
  7. Math. Oper. Und. Stat. Ser. stat. v.11 Some features of the optimal design theory-A survey Pazman, A.
  8. Communications in Statistics v.A7 An Introduction to design optimality with an overview of the literature Ash, A.;Gedayat, A.
  9. Technometrics v.16 An algorithm for construction of D-optimal experimental designs Mitchell, T.J.
  10. Math. Div. Ann. Progr. Rept. Oak Ridge, Tn. Use of design repair to con- struct designs for special linear models Mitchell, T.J.;Miller,F.L.Jr.
  11. Annals of Mathematical Statistics v.41 The sequential generation of D-optimum experimental designs Wynn, H.P.
  12. Theoy of Optimal Experiments Fedorov, V.V.;W.J. Studden(ed.);E.M. Klimko(ed.)
  13. Technometrics v.17 D-optimality for regression designs: A review St.John, R.C.;Draper, N.R.
  14. Technometrics v.22 TIme and space-saving computer methods, related to Mitchell's DETMAX, for finding D-optimum designs Galil, Z.;Kiejer, J.
  15. Annals of Statistics v.8 D-optimum weighing designs Galil, Z.;Kiefer, J.
  16. Technometrics v.25 Some guidelines for constructing exact D-optimal designs on convex design space Johnson, M.E.;Nachtsheim, C.J.
  17. Operation Res. v.21 Comment on integer Maximization probelm Wienstein, I.J.;Yu, S.O,
  18. Technometrics. v.24 Branch and bound search for experimental desings based on D-optimality Welch, W.J.