Hello! I'm a PhD student at UCL working in Bayesian computation, supervised by Dr Sam Livingstone. I'm particularly interested in Adaptivity and Preconditioning in Stochastic Computation.
The occlusion process: improving sampler performance with parallel computation and variational approximation - Hird, M. and Maire, F. (2024). In Submission
Improving the convergence of Markov chains via permutations and projections - Choi, M. C. H., and Hird, M. and Wang, Y. (2024). In Submission
Quantifying the effectiveness of linear preconditioning in Markov chain Monte Carlo - Hird, M. and Livingstone, S. (2023). In Submission
A fresh take on ‘Barker dynamics’ for MCMC - Hird, M., Livingstone, S., and Zanella, G. (2020). in Proceedings of Monte Carlo and Quasi-Monte Carlo Methods 2020
Poster Prize - BAYSM 2024
Deepti Jayawardena Wilkinson Award for Service - Awarded for services to the Department of Statistics at UCL
Poster Prize - London Wide Poster Day
UK-Canada Globalink doctoral exchange scheme - funded by UKRI and Mitacs £15,000 + $3,000 CAD
Departmental Travel Grant to MCQMC c. £1,500
Fully-funded 4 year EPSRC PhD Scholarship c. £70,000 + tuition
Preconditioning for MCMC for Algorithms and Computationally Intensive Inference Seminars (Warwick) 10th February 2023 Slides
Quantifying the effectiveness of linear preconditioning in Markov chain Monte Carlo for MCQMC (Waterloo) 20th August 2024 Paper Slides
Why Boltzmann Brains are Bad for Young Researchers Seminar Series (UCL) 16th November 2022 Slides
Temporal Inference with Finite Factored Sets for Foundations and Landmarks, Year 1 (UCL) 28th June 2021
The MCMC Revolution for Foundations and Landmarks, Year 1 (UCL) 14th December 2020 Slides
I founded and currently organise the Foundations and Landmarks reading group along with Kevin Han Huang and Hugh Dance. We have a website. In year 1 we reviewed foundational papers in probability and statistics. In year 2, and we read Topics in Random Matrix Theory from Terry Tao. It is year 3, we have read Computational Optimal Transport by Peyré and Cuturi, and we are now reading Random Walks on Lie Groups by Emmanuel Breuillard. If you want to join (please do!) send me an email at the address below.
I currently organise the MCMC reading group at the Department of Statistics at UCL. We are reading through Convex Optimization: Algorithms and Complexity from Sebastien Bubeck
I am currently unavailable for consulting work.
The Convergence Group provide insurance to banks against the event that their debtors default. I am building and maintaining their model of losses due to default for a given portfolio of loans.
I partnered with AIXO to create an economic model for the perfume industry. We used the model to simulate market dynamics and infer equilibria. The results are used in their founding document.
Reviewer for AISTATS 2024
Pratyusha Sarkar - Undergraduate Project - Boosting MCMC efficiency with leading eigeninformation
Rui Zhang - Undergraduate Project - Markov Processes and Markov Chain Monte Carlo (Co-supervised with Sam Livingstone)
STAT0023 - Introduction to Practical Statistics - 2nd term 2020/22
My email is max [dot] hird [dot] 19 [at] ucl [dot] ac [dot] uk