Communications for Statistical Applications and Methods

The Korean Statistical Society (한국통계학회)
권호 표지

An Improvement on Estimation for Causal Models of Categorical Variables of Abilities and Task Performance

  • Kim, Sung-Ho (The author wishes to acknowledge the financial support of the Koren Research Foundation made in the program year of 1998)
  • Published : 2000.04.01
Download PDF
⟨ Previous Next ⟩

Abstract

The estimates from an EM when it is applied to a large causal model of 10 or more categorical variables are often subject to the initial values for the estimates. This phenomenon becomes more serious as the model structure becomes more serious as the model structure becomes more complicated involving more variables. In this regard Wu(1983) recommends among others that EMs are implemented several times with different sets of initial values to obtain more appropriate estimates. in this paper a new approach for initial values is proposed. The main idea is that we use initials that are calibrated to data. A simulation result strongly indicates that the calibrated initials give rise to the estimates that are far closer to the true values than the initials that are not calibrated.

Keywords

References

  1. Journal of Royal statistical Society B v.41 no.1 Conditional independence in statistical theory Dawid, A.P.
  2. Journal of Statistical Planning and Inference v.70 Calibration and refinement for classification trees Fienberg, S.E.;Kim, S.H.
  3. Journal of the Korean Statistical Society v.27 no.1 Initial value selection in applying an EM algorithm for recursive models of categotical variables Jeong, M.S.;Kim, S.H.;Jeong, K.M.
  4. Studies in Subjective Probability(edited) Kyberg, Jr., H.E.;Smokler, H.E.
  5. Probabilistic Reasoning in Intelligence Systems: Neworks of Plausible Inference Pearl, J.
  6. The Foundation of Statistics(Second Revised Edition) Savage, L.J.
  7. Bayesian Statistics 4 Learning in probabilistic expert systems Spiegelhalter, D.J.;Cowell, R.G.;Bernardo, J.M.;Berger, J.O.;Dawid, A.P.;Smith, A.F.M.(eds.)
  8. Statistical Science v.8 no.3 Bayesian analysis in expert systems(with discussion) Spiegelhalter, D.J.;Dawid, A.P.;Lauritaen, S.L.;Cowell, R.G.
  9. Technical Report 91-4 Empirical evaluation of prior beliefs about frequencies: methodology and a case study in congenital heart disease Spiegelhalter, D.J.;Harris, N.L.;Bull, K.;Franklin, R.C.G.
  10. Networks v.20 Sequential updating of conditional probabilities on directed graphical structures Spiegelhalter, D.J.;Lauritzen, S.L.
  11. Department of Mathematics and Computer Science Score and information for recursive exponential models with incomplete data. Technical report R-96-2024 Thiesson, B.
  12. Journal of Computational and Graphical Statistics v.6 no.2 On the ordering and groupings of conditional maximizations within ECM-type algorithms Van Dylk, D.;Meng, X.L.
  13. Graphical models in applied multivariate statistics Whittaker, J.
  14. The Annals of Statistics v.11 no.1 On the convergence properties of the EM algorithm Wu, C.F.J.