Interests
- Bayesian Statistics/Machine Learning
- Functional Programming
- Probabilistic Algorithms/Data Structures
- Software/Systems Development
Education
The University of Auckland
PhD, Statistics (2011 - Present)
Memorial University of Newfoundland
MAS, Statistics (2007 - 2010)
BSc, Statistics, Economics (2001 - 2006)
Recent Employment
Senior Developer (2013 - Present)
Bdellium
Senior Research Analyst (2009 - 2011)
Department of Transportation and Works
Government of Newfoundland & Labrador
Research Associate (2007 - 2009)
Department of Fisheries and Oceans | Pêches et Océans Canada
Government of Canada | Gouvernement du Canada
Teaching Assistant (2007 - 2009)
Department of Mathematics & Statistics
Memorial University of Newfoundland
Research & Teaching Assistant (2003 - 2007)
Department of Economics
Memorial University of Newfoundland
Professional Committees
Federal-Provincial-Territorial Committee on Transportation Statistics
Provincial Representative, Newfoundland & Labrador (2009 - 2011)
Statistics Canada | Statistique Canada
Full Cost Investigation of Transportation Working Group
Provincial Representative, Newfoundland & Labrador (2009)
Transport Canada | Transports Canada
Publications & Research
Peer-Reviewed Publications
- Tobin, J. (2010) Approximate marginal inference in models with stratum nuisance parameters, with applications to fishery data. MAS Thesis, Memorial University of Newfoundland. (paper)
- Cadigan, N. and Tobin, J. (2010) Estimating the negative binomial dispersion parameter with highly stratified surveys. Journal of Statistical Planning and Inference 140, 2138-2147. (paper) (poster)
Seminars & Presentations
- (2012) flat-mcmc: Painless general-purpose sampling. Department of Statistics, The University of Auckland, New Zealand. (slides)
- (2012) Bayesian nonparametrics and scalable probabilistic inference. Department of Statistics, The University of Auckland, New Zealand. (slides)
- (2011, Invited) Machine learning for prediction and decision support. Department of Computer Science, Memorial University of Newfoundland, Canada. (slides)
- (2011) Saddlepoint approximations, likelihood asymptotics, and approximate conditional inference. Department of Statistics, The University of Auckland, New Zealand. (slides)
- (2010, with Noel Cadigan) Estimating the negative binomial dispersion parameter with highly stratified surveys. International Statistical Ecology Conference 2010, University of Kent, UK. (poster)
Workshops & Short Courses
- (2012) CEA-EDF-INRIA Summer School in Parallel Functional Programming, Saint Paul lez Durance, France. (github)
- (2011) Machine Learning Summer School 2011, National University of Singapore, Singapore.
- (2010) AD Model Builder Workshop, University of Kent, UK.
- (2007) Advanced R Programming, Meeting of the Statistical Society of Canada, St. John’s, Canada.
Open-Source Projects
- (2012) prompt-pcap: High-throughput financial data parsing.
- (2012) flat-mcmc: Painless general-purpose sampling in Haskell.
- (2012) hasty-hamiltonian: Speedy gradient-based traversal through parameter space.
Some Preferred Tools
Programming
- Haskell
- Unix shell
- C
- R
Computing
- OS (Debian/Ubuntu Linux, Mac OS X)
- Text editing (Vim)
- Quality control (Git, QuickCheck, Travis CI)
- Visualization and graphics (Keynote, d3.js, ggplot2, TikZ)
- Markup (LaTeX/XeTeX, Markdown)
