Theory of Probability and Mathematical Statistics
Closed-Form Estimator for the Matrix-Variate Gamma Distribution
G. Alfelt
Link
Abstract: In this paper we present a novel closed-form estimator for the parameters of the matrix-variate gamma distribution. The estimator relies on the moments of a transformation of the observed matrices, and is compared to the maximum likelihood estimator (MLE) through a simulation study. The study reveals that when the underlying scale matrix parameter is ill-conditioned, or when the shape parameter is close to its lower bound, the suggested estimator outperforms the MLE, in terms of sample estimation error. In addition, since the suggested estimator is closed-form, it does not require numerical optimization as the MLE does, thus needing shorter computation time and is furthermore not subject to start value sensitivity or convergence issues. Finally, regarding the case of general parameter values, using the proposed estimator as start value in the optimization procedure of the MLE is shown to substantially reduce computation time, in comparison to using arbitrary start values.
Keywords: Near-singular matrices, estimation error, maximum likelihood method, asymptotic distribution
Bibliography: 1. G. Alfelt, Stein-Haff identity for the exponential family, Theory of Probability and Mathematical Statistics 99 (2018), no. 2, 7-18.
2. G. Alfelt, T. Bodnar, and J. Tyrcha, Goodness-of-fit tests for centralized Wishart processes, Communications in Statistics - Theory and Methods 49 (2020), no. 20, 5060-5090.
3. S. Anatolyev and N. Kobotaev, Modeling and forecasting realized covariance matrices with accounting for leverage, Econometric Reviews 37 (2018), no. 2, 114-139.
4. T. W. Anderson, An introduction to multivariate statistical analysis, Wiley, 2003.
5. V. Golosnoy, B. Gribisch, and R. Liesenfeld, The conditional autoregressive Wishart model for multivariate stock market volatility, Journal of Econometrics 167 (2012), no. 1, 211-223.
6. G.H. Golub and C.F. Van Loan, Matrix computations, Johns Hopkins University Press, Baltimore, 2013.
7. A. K. Gupta and D. K. Nagar, Matrix variate distributions, CRC Press, 2000.
8. A. Iranmanesh, M. Arashi, D. K. Nagar, and S. M. M. Tabatabaey, On inverted matrix variate gamma distribution, Communications in Statistics - Theory and Methods 42 (2013), no. 1, 28-41.
9. R. J. Muirhead, Aspects of multivariate statistical theory, Wiley, New York, 1982.
10. R. Vani Lakshmi and V.S. Vaidyanathan, Parameter estimation in multivariate gamma distribution, Statistics, Optimization and Information Computing 3 (2015), 147-159.
11. Z. S. Ye and N. Chen, Closed-form estimators for the gamma distribution derived from likelihood equations, The American Statistician 71 (2017), no. 2, 177-181.