# Dr Krzysztof Łatuszyński

**Royal Society University Research Fellow (2014-2022)**

**I'm currently highlighting the following**

**Applications welcome: Research Associate**funded by the Royal Society to work with me on high dimensional MCMC. Deadline 17th of November. Details here.

**seminar: Algorithms & Computationally Intensive Inference****Simulation Reading Group**- I gave a short course on
**Adaptive MCMC**at**XIV EBEB**in Rio.**Here are the slides**. - My lectures on
**MCMC for intractable likelihood problems**from the Mathematical Statistics conference in Bedlewo are here.

#### Research Interests

**general:**stochastic simulation, computational statistics, Bayesian statistics**specific:**probabilistic algorithms, MCMC, adaptive MCMC, perfect simulation, analysis of algorithms, exact inference for SDEs, intractable likelihood problems- I also like: applied statistics, algebraic statistics, optimal transport, combinatorics...

#### Teaching

- In 2018/2019 I'm teaching OxWaSP module 1 on statistical computing with Robin Evans.

See the course webpage here.

#### Ph.D. Students

- Timothee Stumpf-Fetizon - co-supervised with Gareth Roberts
- Giulio Morina (OxWaSP) - co-supervised with Ewan Cameron
- Emilia Pompe (OxWaSP) - co-supervised with Chris Holmes

[see: paper 1] - Cyril Chimisov - co-supervised with Gareth Roberts

graduated 2018, now at Google

[see: paper 1, paper 2, paper 3]

Papers and preprints (theory and methodology)

for latest contributions see my google scholar

2019

**From the Bernoulli Factory to a Dice Enterprise via Perfect Sampling of Markov Chains**(with Giulio Molina and Alex Wendland) [arxiv:]

2018

**A Framework for Adaptive MCMC Targeting Multimodal Distributions**(with Emilia Pompe and Chris C. Holmes) [arxiv:1812.02609]

(supersedes this note pdf)

The Annals of Statistics (to appear)**Air Markov Chain Monte Carlo**(with Cyril Chimisov and Gareth O. Roberts) [arxiv:1801.09309]

**Adapting The Gibbs Sampler**(with Cyril Chimisov and Gareth O. Roberts) [arxiv:1801.09299]

2017

**Continuous-time Importance Sampling: Monte Carlo Methods which Avoid Time-discretisation Error**(with Paul Fearnhead, Gareth O. Roberts and G. Sermaidis) [arxiv:1712.06201]

**Exact Monte Carlo likelihood-based inference for jump-diffusion processes**

(with Flavio B. Gonçalves, Gareth O. Roberts) [arxiv:1707.00332]**Barker’s algorithm for Bayesian inference with intractable likelihoods**

(with Flavio B. Gonçalves, Gareth O. Roberts) [arxiv: 1709.07710]

*Brazilian Journal of Probability and Statistics*31(4), pp. 732-745, (2017)**In Search of Lost (Mixing) Time: Adaptive Markov chain Monte Carlo schemes for Bayesian variable selection with very large p**

(with Jim Griffin and Mark Steel) [arxiv: 1708.05678]

2015

**Bayesian computation: a summary of the current state, and samples backwards and forwards**(with P. Green, M. Pereyra, C. P. Robert) [arxiv:1502.01148]

*Statistics and Computing, 2015*, 25(4), pp. 835-862, (2015)

2014

**Individual adaptation: an adaptive MCMC scheme for variable selection problems**(with Jim Griffin and Mark Steel) [arXiv:1412.6760v2]

Matlab code is available from Jim's webpage**Convergence of hybrid slice sampling via spectral gap**(with Daniel Rudolf) [arxiv:1409.2709]

**Perfect simulation using atomic regeneration with application to Sequential Monte Carlo**(Anthony Lee, Arnaud Doucet, Krzysztof Łatuszyński) [arxiv:1407.5770]

**Stability of Adversarial Markov Chains, with an Application to Adaptive MCMC Algorithms**(with Radu V. Craiu, Larry Gray, Neal Madras, Gareth O. Roberts and Jeffrey S. Rosenthal) [arxiv:1403.3950]

*The Annals of Applied Probability*25(6), pp. 3592-3623, (2015)

2013

**The Containment Condition and AdapFail algorithms**

(with Jeffrey S. Rosenthal) [arxiv:1307.1799]

*Journal of Applied Probability,*51(4), pp. 1189-1195, (2014)

2012

**Variance bounding and geometric ergodicity of Markov chain Monte Carlo kernels for approximate Bayesian computation**(with Anthony Lee) [arxiv:1210.6703]

*Biometrika*101(3), pp. 655-671, (2014)

2011

**Nonasymptotic bounds on the****estimation error of MCMC algorithms**(with Blazej Miasojedow and Wojciech Niemiro)

*Bernoulli,*19(5A), pp. 20133-2066, (2013)**CLTs and asymptotic variance of time sampled Markov chains**

(with Gareth O. Roberts) [arxiv:1102.2171]

*Methodology and Computing in Applied Probability*, 15(1), pp. 237-247, (2013)**Adaptive Gibbs samplers and related MCMC methods**(with Gareth O. Roberts and Jeffrey S. Rosenthal) [.pdf / arxiv:1101.5838]

(Major revision of an earlier two-author paper)

*The Annals of Applied Probability*, 23(1), pp. 66-98, (2013)

2010

**Nonasymptotic bounds on the mean square error for MCMC estimates via renewal techniques**(with Blazej Miasojedow and Wojciech Niemiro) [arxiv:1101.5837]

*MCQMC 2010 Conference Proceedings*(2012)**Adaptive Gibbs samplers**(later revised into arxiv:1101.5838)

(with Jeffrey S. Rosenthal) [arxiv:1001.2797]

2009

**Rigorous confidence bounds for MCMC under a geometric drift condition**

(with Wojciech Niemiro) [arxiv:0908.2098]

*Journal of Complexity,*27(1), pp. 23–38, (2011)**Nonasymptotic bounds on the estimation error for regenerative MCMC algorithms**(later revised into arxiv:1106.4739 and arxiv:1101.5837)

(with Blazej Miasojedow and Wojciech Niemiro) [arxiv:0907.4915]**Simulating Events of Unknown Probabilities via Reverse Time Martingales**

(with Ioannis Kosmidis, Omiros Papaspiliopoulos, Gareth O. Roberts) [arxiv:0907.4018]

*Random Structures and Algorithms,*38(4), pp. 441-452, (2011)

2008

**A Regenerative Proof of the Central Limit Theorem for Uniformly Ergodic Markov Chains**

(with Witold Bednorz and Rafal Latala)

*Electronic Communications in Probability*, 13, pp. 85-98, (2008)

2007

**A few Remarks on "Fixed-Width Output Analysis for Markov Chain Monte Carlo" by Jones et al.**

(with Witold Bednorz) [.pdf]

*Journal of the American Statistical Association,*102(480), pp. 1485-1486, (2007)

##### other research contributions

(PhD Thesis, conference proceedings, book chapters & contributions, discussions, reports, etc...,

... and papers OTHER THAN theory and methodology)

2014

**Discussion of "Sequential Quasi-Monte-Carlo Sampling" by M. Gerber and N. Chopin**

(M. Pollock, A. M. Johansen, K. Latuszynski and G. O. Roberts) [.pdf]

*Journal of the Royal Statistical Society B*, (In Press)**P1042: Relation of sleep stages and sleep period to the cortical excitability in Parkinson’s disease**

(J Antczak, M Rakowicz, K Latuszynski, J Phelps, A Sobanska, E Inglot, U Zalewska, T Jakubczyk)

Clinical Neurophysiology 125, S326

2011

**The Differences in Sleep Profile Changes under Continous****Positive Airway Pressure (CPAP) Therapy between Nonobese, Obese and****Severely Obese Sleep Apnea Patients**

(J. Antczak, B. Horn, A. Richter, R. Bodenschatz, K. Latuszynski, E.W. Schmidt, W. Jernajczyk)

*Journal of Physiology and Pharmacology*, 63(4), pp. 263-269, (June 2012)

2010

**Discussion of Girolami M. and Calderhead B.**(with Gareth O. Roberts, Alexandre Thiéry and Katarzyna Wolny)

*Journal of the Royal Statistical Society B*, 73(2), pp. 188-189, (2011).**Discussion of Andrieu C., Doucet A. and Holenstein R.**(with Omiros Papaspiliopoulos)

*Journal of the Royal Statistical Society B*, 72(3), pp. 310-311, (2010).**The Bernoulli Factory, its extensions and applications**[.pdf]International Workshop on Applied Probability - IWAP2010 Short paper

2009

**Optimal transports and c-convex functions**

*Oberwolfach Reports*, Volume 6, Issue 2, 2009,*Arbeitsgemeinschaft: Optimal Transport and Geometry*organised by Felix Otto and Karl-Theodor Sturm**Quasi-symmetry and Cycles**

(with Carlos Trenado)

I had the pleasure to contribute (pp. 128-131) to the problem section of*Lectures on Algebraic Statistics*by M. Drton, B. Sturmfels and S. Sullivant, Series: Obervolfach Seminars, Vol.39, Birkhäuser 2009

2008

**Wage Differences between Woman and Man in Polish Labor Market in 2004**

(with Lukasz Wozny)

a book chapter in*Wzrost gospodarczy a bezrobocie i nierownosci w podziale dochodu,*ed. Marek Garbicz, Wojciech Pacho, SGH, Warszawa 2008**Regeneration and Fixed-Width Analysis of Markov Chain Monte Carlo Algorithms**

[arxiv:0907.4716]

**PhD Thesis**(2008), University of Warsaw, supervisor: Wojciech Niemiro

2007

**A Regeneration Proof of the Central Limit Theorem for Uniformly Ergodic Markov Chains**

(with Witold Bednorz)

*15th European Young Statisticians Meeting - EYSM Conference Proceedings, 2007.*(see the substantially extended 2008 ECP paper!)

2006

**MCMC-Approximation under Drift Condition**

(with Wojciech Niemiro)

*Proceedings of the 6th International Workshop on Rare Event Simulation (RESIM 2006).*(see also the improved and extended 2009 paper and Chapter 5 of my PhD Thesis)

Mathematical Sciences Building

Rm 3.03, Dept of Statistics

University of Warwick

Coventry, CV4 7AL

Tel: 02476 150920

Fax: 02476 524532

**Email:**

K.G.Latuszynski 'at' warwick.ac.uk

- Royal Society University Research Fellow (2014-2022)

- Facuty Fellow, The Alan Turing Institute

I am also a member of the WDSI

Prior to that I did postdocs/RFs in CRISM (2008-09), at Toronto (2009) and Warwick (2009-12) and was Jeff Harrison Assistant Professor (2012-16)