Associate ProfessorProf PJ (Paul) Mostert
Associate Professor
University of Stellenbosch
Private Bag X1, Matieland
7602, South Africa
Tel: +(27) 21 808 3536
Fax: +(27) 21 808 3830
e-mail: pjmos@sun.ac.za
Born in Klerksdorp. North West Province, South Africa on 29 March 1962. Obtained the B.Sc, HED and B.Sc(Honors) in Statistics at Potchefstroom University for Christian Higher Education. The M.Sc in Mathematical Statistics was obtained at the University of Pretoria.
Obtained a Ph.D in 1999 at UNISA in the Department of Statistics. Title of thesis: "A Bayesian method to analyse cancer lifetimes using Rayleigh models", with promoters professors Kotie Roux and Andriëtte Bekker.
Biostatistics I, Mathematical Statistics II, Statistics II, Calculating Interest I
Post Graduate Course: Bayesian Statistics (Honors and Masters students)
Repairable systems and their evaluation of their performance in terms of reliability and availability. A Bayesian method of assessing reliability is of primary interest, since very little is published on this topic. I registered this research project with the NRF and submitted an application for a research grant (2003 – 2006) in April 2002. The title of the project is: "System Reliability Analysis" within the focus area of "Information and Communication Technology".
Co-workers: Professor Sarma Yadavali (University of Pretoria)
Professor Andriette Bekker (UNISA)
Dr Magda Botha (UNISA)
Bayesian shrinkage estimators for cancer lifetime models. Empirical Bayes shrinkage estimators for the lifetime parameters are studied under different loss functions, when cancer lifetimes are observed. Due to knowledge of cancer patients in the past one may have a guessed or prior value of a measure for the middle, which may be used in constructing an estimate of the lifetime parameter. This is a new project and started in 2002.
Co-workers: Professor JJJ (Kotie) Roux (Research Fellow, UNISA)
Professor Andriette Bekker (UNISA)
Process prior methodology within the Cox regression models and non-parametric Bayesian survival analysis are investigated. Estimation of the baseline hazard and cumulative baseline hazard with or without ties are the main focus. Estimation is done using the Bayesian approach. With Mr Chris Muller (Department of Statistics and Actuarial Science, Stellenbosch)
1986 - 1987 Teacher - Potchefstroom Gymnasium
1988 - 1988 Actuarial Assistant - Momentum Life
1988 - 1994 Operations Researcher and Statistician - ISCOR
1994 - 1997 Lecturer - UNISA
1994 - 1997 Part-time Senior Lecturer - MEDUNSA
Mostert, P.J., Bekker, A. and Roux, J.J.J. (1998). Bayesian analysis of survival data using the Rayleigh model and linex loss. South African Statistical Journal. Volume 32, No 1, p. 19-42.
Mostert, P.J., Roux, J.J.J. and Bekker, A. (1999). Bayes estimators of the lifetime parameters using the compound Rayleigh model. South African Statistical Journal, Volume 33, No 2, p 117-138.
Bekker, A., Roux, J.J.J and Mostert, P.J. (2000). A generalization of the compound Rayleigh distribution: Using Bayesian methods on cancer survival times. Communications of Statistics – Theory and Methods, Volume 29, No 7, p1419-1433.
Yadavalli, V.S.S., Bekker A., Mostert, P.J. and Botha, M. (2001). Bayesian estimation of the stationary rate of disappointments for a model of a two-unit intermittently used system. Pakistan Journal of Statistics, Volume 17, No 2, p117-125.
Yadavalli, V.S.S., Bekker, A. Mostert, P.J. & Botha (2001). Bayesian highest posterior density intervals for the availability of a system with 'rest-period' for the repair facility. South African Journal of Industrial Engineering, Volume12, No 2, p17-24.
Yadavalli, V.S.S., Mostert, P.J., Bekker, A. and Botha, M. (2002). Computation of posterior distribution in Bayesian analysis: application in an intermittently used reliability system ("Berekening van 'n aposteriori-verdeling in Bayes-analise: toepassing in 'n betroubaarheidstelsel wat afwisselend gebruik word"). Die Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Volume 21(3), p. 78-82.
Mostert, P.J., Bekker, A. & Yadavalli, V.S.S. (2004). A two-unit intermittently used cold standby system from a Bayesian viewpoint assuming a Dirichlet prior. Pakistan Journal of Statistics, Volume 20, No 2, p. 223-229.
Mostert, P.J., Bekker, A. & Roux, J.J.J. (2007). Shrinkage estimation of the lifetime parameters in the Rayleigh model using an empirical Bayesian approach. International Journal of Statistics and Management Systems. Volume 1(1-2), p. 143-159.
Mostert, P.J., Bekker, A. and Roux, J.J.J. (2001). A Bayesian method to analyse cancer survival times using the Weibull model. Proceedings of ISBA 2000 – Bayesian Methods with applications to Science, policy and Official Statistics. Selected papers from ISBA 2000: The sixth world Meeting of the International Society for Bayesian Analysis, p371-380. (Refereed).
Mostert, P.J., Bekker, A., Yadavalli, V.S.S. and Botha, M. (2001). The study of the posterior distribution of the stationary rate of disappointments for an intermittently used two-unit cold standby system. Proceedings of the Xth International Symposium on 'Applied Stochastic Models and Data Analysis', Volume 2 p767-772. (refereed).
Bekker, A., Roux, J.J.J. and Mostert, P.J. (1993). Bayesian estimation for the scale parameter of the Rayleigh distribution with censoring. Research report 93/5, Department of Statistics, University of South Africa.
Mostert, P.J., Bekker, A. and Roux, J.J.J. (1994). A comparative study of empirical Bayesian shrinkage estimation of the scale parameter of the censored Rayleigh distribution. Research report 94/2, Department of Statistics, University of South Africa.
Roux, J.J.J., Mostert, P.J. and Bekker, A. (1994). A Bayes analysis of the censored Rayleigh model when using a generalized hypergeometric as a prior. Research report 94/3, Department of Statistics, University of South Africa.
Bekker, A., Roux, J.J.J. and Mostert, P.J. (1995). Bayesian reliability analysis for a two-parameter Rayleigh distribution using an asymmetric linex loss function. Research report 95/4, Department of Statistics, University of South Africa.
Mostert, P.J., Bekker, A., and Roux, J.J.J. (1997). Bayesian estimation of some lifetime distribution identities of the Rayleigh model using linex loss function. Research report 97/1, Department of Statistics, University of South Africa.
Mostert, P.J., Bekker, A., and Roux, J.J.J. (1998). A Bayesian method to analyse cancer lifetimes using the Weibull model. Research report 98/3, Department of Statistics, University of South Africa.
Mostert, P.J., Roux, J.J.J. and Bekker, A. (1999). Bayes estimators of the lifetime parameters using the compound Rayleigh model. Research report 99/3, Department of Statistics, University of South Africa.
Yadavalli, V.S.S., Bekker A., Mostert, P.J. and Botha, M. (2000). Bayesian estimation of the stationary rate of disappointments for a model of a two-unit intermittently used system. Research report 00/4, Department of Statistics, University of South Africa.
Mostert, P.J., Bekker, Yadavalli, V.S.S. and Botha, M. (2000). Bayes estimation of stationary rate of disappointments for two models of an intermittently used system. Research report 00/7, Department of Statistics, University of South Africa.
Jordaan, B., Griesel, R.D. and Mostert, P.J. (1995). Comparison of three different anti-epileptic formulations on the bio-availability and electro-encephalographic parameters in healthy volunteers after multiple dosing. International Neuro-psychology Satellite Symposium, Hayman Island, Barrier Reef, Australia. June 1995.
Mostert, P.J., Roux, J.J.J. and Bekker, A. (1997). Bayesian estimation of some lifetime distribution identities of the Rayleigh model using linex loss function. 51st Session of the International Statistical Institute, Istanbul. August 1997.
Bekker, A., Roux J.J.J. and Mostert, P.J. (1999). A Bayesian method to analyse cancer lifetimes using the compound Weibull model. 52nd Session of the International Statistical Institute, Helsinki. August 1999.
Mostert, P.J., Roux, J.J.J. and Bekker, A. (2000). A Bayesian method to analyse cancer lifetimes using the Weibull model. 6th World Meeting of the International Society for Bayesian Analysis, Heraklion, Hersonissos, Crete, Greece, July 2000.
Mostert, P.J., Bekker, A., Yadavalli, V.S.S. and Botha, M. (2001). The study of the posterior distribution of the stationary rate of disappointments for an intermittently used two-unit cold standby system. The Xth International Symposium on APPLIED STOCHASTIC MODELS AND DATA ANALYSIS, Compiégne, France, June 2001.
Mostert, P.J. and Bekker, A. (2002). A two unit intermittently used cold standby system from a Bayesian viewpoint assuming a Dirichlet prior. Seventh Valencia International Meeting on Bayesian Statistics, Tenerife, Canary Islands, Spain, June 2002.
Bekker, A. and Mostert, P.J. (2002). Bayesian inference of the steady state availability of a one-unit repairable system with 'rest-period'. Seventh Valencia International Meeting on Bayesian Statistics, Tenerife, Canary Islands, Spain, June 2002.
Mostert, P.J. (2004). Segmented estimation of the hazard rate estimation for the two parameter Weibull model using LINEX loss. ISBA 2004 World Meeting, Vińa del Mar, Chile, May 2004.
Mostert, P.J., Bekker, A. and Roux, J.J.J. (2006). Shrinkage estimation of lifetime parameters in the Rayleigh model using an empirical Bayesian approach. Eighth Valencia International Meeting on Bayesian Statistics/ISBA2006, Benidorm, Spain, June 2006.
Muller, C.J.B. and Mostert, P.J. (2006). Non-parametric Bayesian multistate modelling of paediatric HIV. Eighth Valencia International Meeting on Bayesian Statistics/ISBA2006, Benidorm, Spain, June 2006.
Mostert, P.J. 2006. Changing approaches and perceptions: Biostatistics and its role in teaching the Stellenbosch doctor. Seventh International Conference on the Teaching of Statistics (ICOTS 7), Salvador, Brazil, July 2006.
Mostert, P.J. and Du Toit, S.H.C. (1991). Stochastic parameter non-linear regression. Annual conference of the South African Statistical Association, Pretoria. November 1991.
Bekker, A., Roux J.J.J. and Mostert, P.J. (1993). Bayes estimation for the scale parameter of the Rayleigh distribution with censoring. Conference of the South African Statistical Association, Pietermaritzburg. November 1993.
Bekker, A., Roux J.J.J. and Mostert, P.J. (1998). A generalisation of the compound Rayleigh distribution: Using Bayesian methods on survival data. Annual conference of the South African Statistical Association, Midrand. November 1998.
Mostert, P.J., Roux, J.J.J. and Bekker, A. (1998). The type I Rayleigh distribution as a failure model from a Bayesian approach. Annual conference of the South African Statistical Association, Midrand. November 1998.
Mostert, P.J., Roux, J.J.J. and Bekker, A. (1999). A Bayesian method to analyse cancer survival times using the Weibull model. Annual conference of the South African Statistical Association, University of Natal, Durban. November 1999.
Mostert, P.J., Bekker, Yadavalli, V.S.S. and Botha, M. (2000). Bayes estimation of stationary rate of disappointments for two models of an intermittently used system. Annual conference of the South African Statistical Association, University of the Witwatersrand, Johannesburg. November 2000.
Bekker, A., Mostert, P.J., Yadavalli, V.S.S. and Botha, M. (2000). Bayesian estimation of the stationary rate of disappointments for a model of a two-unit intermittently used system. Annual conference of the South African Statistical Association, University of the Witwatersrand, Johannesburg. November 2000.
Bekker A., Mostert, P.J., Yadavalli, V.S.S. and Botha, M. (2001). Bayesian highest posterior density intervals for the availability of a system with rest-period for the repait facility. Annual conference of the South African Statistical Association, University of Cape Town, Goudini Spa. November 2001.
Mostert, P.J., Bekker, A., Yadavalli, V.S.S. and Botha, M (2001). A one-unit intermittently used system from a Bayesian viewpoint assuming conjugate priors. Annual conference of the South African Statistical Association, University of Cape Town, Goudini Spa. November 2001.
Mostert, P.J., Roux, J.J.J. and Bekker. (2003). An empirical Bayes shrinkage estimation of the lifetime parameters in the Rayleigh model. Annual conference of the South African Statistical Association, Caesers Gauteng. November 2003.
Mostert, P.J. (2004). Slice sampler in action: reference posterior analysis for a Weibull model. Annual conference of the South African Statistical Association, University of the Free State, Bloemfontein. November 2004.
Muller, C.J.B. and Mostert, P.J. (2004). Non-parametric Bayesian survival analysis: The Gamma process prior for the baseline cumulative hazard function. Annual conference of the South African Statistical Association, University of the Free State, Bloemfontein. November 2004.
Nagel, C., de Wet, T. and Mostert, P.J. (2005). Bayesian extreme value modeling using non-informative priors. Annual conference of the South African Statistical Association, Rhodes University, Grahamstown. November 2005.
Muller, C.J.B. and Mostert, P.J. (2007). A Bayesian approach to model a multi-state Markov model from interval-censored data. Annual conference of the South African Statistical Association, University of the Witwatersrand, Misty Hills. November 2007.
Member of the South African Statistical Association, 1991-.
Member of the International Society for Bayesian Analysis, 1998-.
Member of the Southern Africa International Society for Bayesian Analysis Executive Committee, 1999-.
Chairperson of the International Society for Bayesian Analysis – Southern Africa Chapter, 2000-.
Member of the Executive Committee of the South African Statistical association, 2005-.
Rated by the National Research Foundation (NRF) in the Y category: August 2001- December 2006.