Classical and Bayesian prediction for multivariate simple regression model
Paper
Paper/Presentation Title | Classical and Bayesian prediction for multivariate simple |
---|---|
Presentation Type | Paper |
Authors | |
Author | Khan, Shahjahan |
Editors | Ahmed, Munir |
Journal or Proceedings Title | Journal of Applied Probability and Statistics |
Journal Citation | 9 (1), pp. 67- 78 |
Number of Pages | 12 |
Year | 2014 |
Publisher | Islamic Countries Society of Statistical Sciences (ISOSS) |
Place of Publication | Lahore, Pakistan |
ISSN | 1930-6792 |
ISBN | 9789698858162 |
Web Address (URL) of Paper | http://japs.isoss.net/0803.pdf |
Conference/Event | 13th Islamic Countries Conference on Statistical Sciences |
Event Details | 13th Islamic Countries Conference on Statistical Sciences Parent Islamic Countries Conference on Statistical Sciences (ICCS) Delivery In person Event Date 18 to end of 21 Dec 2014 Event Location Bogor, Indonesia |
Abstract | Both Bayesian and classical approaches are used to derive the prediction distribution of a set of future responses, conditional on another set of independent realized responses, from the multivariate simple regression model in this paper. The errors from both the performed and future experiments are assumed to be identically and independently distributed as multivariate normal variables. Conditional on the realized responses, the future unrealized responses follow a matrix T distribution. The shape parameter of the prediction distribution depends on the size of the realized sample, and the dimension of the regression parameters in the model. The prediction distribution obtained by both the classical method and Bayesian method under uniform prior is the same. |
Keywords | multivariate normal and Student-t distributions; matrixnormal, matrix gamma and matrix T distributions; matrix integration; invariant differentials; uniform prior; and prediction distribution. |
ANZSRC Field of Research 2020 | 490599. Statistics not elsewhere classified |
490509. Statistical theory | |
490506. Probability theory | |
Public Notes | Full Reference: Khan, S (2014). Classical and Bayesian prediction for multivariate simple regression model. Journal of Applied Probability and Statistics. Vol. 9(1), 67-78 |
Byline Affiliations | Australian Centre for Sustainable Catchments |
Institution of Origin | University of Southern Queensland |
https://research.usq.edu.au/item/q2w5x/classical-and-bayesian-prediction-for-multivariate-simple-regression-model
1664
total views7
total downloads1
views this month0
downloads this month