Estimation and Optimal Designing under Latent Variable Models for Paired Comparisons Studies via a Multiplicative Algorithm
1. The problem of estimating the parameters of latent variable models such as the Bradley Terry or Thurstone Model by the method of maximum likelihood, given data from a paired comparisons experiment. The parameters of these models can be taken to be weights which are positive and sum to one;
2. The problem of determining approximate locally optimal designs for good estimation of these parameters; i.e of determining optimal design weights which are also positive and sum to one.
KeywordsOptimal Design Paired Comparison Latent Variable Model Thurstone Model Multiplicative Algorithm
Unable to display preview. Download preview PDF.
- Torsney, B. (1988) Computing Optimizing Distributions with Applications in Design, Estimation and Image Processing. In Optimal Design and Analysis of Experiments (Y. Dodge, V.V. Fedorov and H.P. Wynn eds). Amsterdam: North Holland, 361-370.Google Scholar
- Torsney, B. (2004) Fitting Bradley Terry models using a multiplicative algorithm. In Proceedings in Computational Statistics (COMPSTAT2004) (J. Antoch, ed.).Heidelberg: Physica-Verlag, 214-226.Google Scholar
- Torsney, B. and Alahmadi, A. M. (1992) Further developments of algorithms for constructing optimizing distributions. In Model Oriented Data Analysis (Proceedings of 2nd IIASA- Workshop, St. Kyrik, Bulgaria, 1990) (V. Fedorov, W.G. Muller, I.N. Vuchkov eds). Heidelberg: Physica-Verlag, 121-129.Google Scholar
- Whittle, P. (1973). Some general points in the theory of optimal experimental design Journal of the Royal Statistical Society B 35, 123-130.Google Scholar