Qualifications:
BSc Honours Degree in Statistics, University of Peradeniya, Sri Lanka
PhD, Queensland University of Technology: “Improving Species Distribution Modelling: Selecting Absences and Eliciting Variable Usefulness for Input into Standard Algorithms or a Bayesian Hierarchical Meta-factor Model”
Research Interests
- Bayesian methods for Ecology
- Bayesian informative priors
- Variable selection methods
- Species Distribution Modelling
- Expert elicitation
Project Description
In species distribution models (SDM), the number of potential variables available in spatial datasets, imagery and acoustic sensing are increasing rapidly. With this increase a problem which remains is a structured framework for deciding which environmental variables to be included. Appropriate selection of environmental variables may greatly influence the performance of the model and its potential application for prediction or explanation. My PhD is focusing on the use of experts’ knowledge by formulating their mathematical representation in appropriate Bayesian informative priors to facilitate variable selection in species distribution models of rare, threatened or pest species.
Conference
J Ramethaa, P Wijekoon, C Santiapillai, S Wijeyamohan. An Assessment of the Coexistence of Humans and Elephants in Sri Lanka. In Proceedings of the Peradeniya University Research Sessions (PURSE), Sri Lanka, Vol. 16, 24th November 2011. URL http://www.pdn.ac.lk/purse/Proceedings/2011/PURSE%202011%20Abstracts.pdf
Pirathiban, R., Low-Choy, S.J., Williams, C., and Murray, J. Effect of visual stimuli (GIS vs photo) on scenario-based elicitation of probabilities of site occupancy. In Proceedings of the 10th International Workshop for the Australasian Chapter of the International Society for Bayesian Analysis (ISBA) and the annual meeting of the Bayesian Statistics section of the Statistical Society of Australia, Inc. (SSAI) – Bayes on the Beach, Gold Coast, Australia 10-12 November 2014. URL https://botb2014.files.wordpress.com/2014/11/conference-program-bob-20141.pdf
Pirathiban, R., Williams, K.J., Pettitt, A.N. and Low-Choy, S.J. Eliciting and encoding expert knowledge on variable selection into classical or Bayesian species distribution models . In Proceedings of the International Biometric Society Australasian Region Conference 2015 – BIOMETRICS by the Harbour, Hobart, Tasmania 30 November – 3 December 2015. URL http://www.biometricsociety.org.au/conferences/Hobart2015/talks2015/Wednesday/W_1430_Wed_RamethaaPirathiban.pdf
Pirathiban, R., Williams, K.J. and Low-Choy, S.J. (2015). Delineating environmental envelopes to improve mapping of species distributions, via a hurdle model with CART &/or MaxEnt. In Weber, T., McPhee, M.J. and Anderssen, R.S. (eds) MODSIM2015, 21st International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, 29 Nov to 4 Dec 2015, Gold Coast, Australia, pp. 1455–1461. ISBN: 978-0-9872143-5-5. www.mssanz.org.au/modsim2015/B4/walmsley.pdf
Pirathiban, R., Williams, K.J., Pettitt, A.N. and Low-Choy, S.J. Eliciting and encoding expert knowledge on variable selection for Species Distribution Models. In Proceedings of the 11th International Workshop for the Australasian Chapter of the International Society for Bayesian Analysis (ISBA) and the annual meeting of the Bayesian Statistics section of the Statistical Society of Australia, Inc. (SSAI) – Bayes on the Beach, Gold Coast, Australia 7-9 December 2015. URL https://botb2015.files.wordpress.com/2015/07/conference-program-bob-2015-final4.pdf
Contact details
- j.ramethaa@qut.edu.au
- ramethaa.jeyapalan@hdr.student.qut.edu.au
- jramethaa@gmail.com
Bayesian Research & Applications Group 