Andi Wang's Personal Page
Email: a dot wang at stats.ox.ac.uk
I am a DPhil student in the Department of Statistics at the University of Oxford. My supervisors are David Steinsaltz (Oxford) and Gareth Roberts (Warwick).
My current research focuses on the theoretical underpinnings of quasistationary Monte Carlo methods. Quasistationary Monte Carlo methods were first proposed in the ScaLE Algorithm, where it was demonstrated that such methods can efficiently perform Bayesian inference in settings with large numbers of data points.
I completed my undergraduate and master's in Mathematics at the University of Cambridge, 2011-2015, graduating with a first class in each year and Distinction in Part III. I specialised in Statistics, Probability and Analysis, and wrote an essay titled Nonparametric Inference Under Shape Constraints. I received Horne Scholarships, the Hockin Prize and Wright Prize, and in Part III I was awarded the Thomas Bond Sprague Prize.
In September 2015 I moved to Oxford to start on the OxWaSP CDT, as part of the 2015-2016 cohort. In my first research project I began to study the theory of quasistationarity and the ScaLE Algorithm, and in my second project I looked at the application of piecewise-deterministic Markov processes to simulation, in particular the Bouncy Particle Sampler. Since September 2016 I have been working with David Steinsaltz and Gareth Roberts on the theoretical properties of quasistationary Monte Carlo methods.
- LMS Computational Statistics Summer School, Warwick University, July 2018. Short talk + poster.
- MCQMC, Rennes, France, July 2018.
- Quasistationary Distributions: Analysis and Simulation, Paderborn University, Germany, September 2017.
- Durham Symposium on Markov Processes, Mixing Times and Cutoff, July 2017. Talk Foundational Properties of Quasistationary Monte Carlo Methods, talk.
- Scalable Statistical Inference workshop, Isaan Newton Institute, Cambridge, July 2017. Poster.
Some of my work
- Wang, A.Q., Kolb, M., Roberts, G.O., Steinsaltz, D., 2017, Theoretical Properties of Quasistationary Monte Carlo Methods, arXiv 1707.08036, submitted.
- Talk: Quasistationary Monte Carlo Methods, April 2018, Fintech Doctoral Seminar Series, Stanford University.
- Poster: Quasistationary Monte Carlo Methods, March 2018, 2nd Year poster session, voted joint best poster.
- Talk: Quasistationary Monte Carlo Methods, Februrary 2018, Algorithms & Computationally Intensive Inference Seminar Series, University of Warwick.
- Poster: Foundational Properties of Quasistationary Monte Carlo Methods, June 2017, Oxford Department of Statistics 1st Year poster session, voted joint best poster.
- Talk: Nonparametric Estimation Under Shape Constraints, June 2015, Part III seminar, University of Cambridge. Video: coming soon.
- Nonparametric Inference Under Shape Constraints, 2015, Part III essay. Supervised by Richard Samworth.