BOSTON UNIVERSITY

Recent Technical Reports


• Data-driven rainfall prediction at a regional scale: a case study with Ghana. Joint with Indrajit Kalita, and Lucia Vilallonga. See also the companion website.
• Laplace approximation for Bayesian variable selection via Le Cam’s one-step procedure. Joint with Tianrui Hou, and Liwei Wang.
• On the estimation rate of Bayesian physics-informed neural networks for inverse problems. Joint with Yi Sun, and Debarghya Mukherjee.
• Unbiased Markov Chain Monte Carlo: what, why, and how. Joint with Pierre Jacob. See also the companion website
• A statistical perspective on unrolling models for inverse problems. Joint with Xinru Liu, and Qiuyun Zhu.
• On generative energy-based models. Joint with Keer Jiang, and Yi Sun.
• An approach to large-scale quasi-Bayesian inference with spike-and-slab prior. Joint with Anwesha Bhattacharyya.
• Bayesian variable selection in linear regression models with instrumental variables.. Joint with Gautam Sabnis and Prosper Dovonon

Publications


• On cyclical MCMC. Joint with Liwei Wang, Xinru Liu, and Aaron Smith. AISTATS 2024.
• Efficiency bounds for moment condition models with mixed identification strength. Joint with Prosper Dovonon, and Firmin D Tchatoka. Journal of Econometrics (2024+)
• Minimax quasi-Bayesian estimation in sparse canonical correlation analysis via a Rayleigh quotient function. Joint with Qiuyun Zhu. JASA (2023+). A supplemental document is available here.
• Probabilistic Unrolling: Scalable, Inverse-Free Maximum Likelihood Estimation for Latent Gaussian Models. Joint with Alexander Lin, Bahareh Tolooshams, and Demba Ba. ICML 2023.
• A fast asynchronous MCMC sampler for sparse Bayesian inference. Joint with Liwei Wang. JRSS-B 85, pages 1492-1516 (2023). Supplemental material here.
• Approximate spectral Gaps for Markov chains mixing times in high dimensions. SIAM Journal on Mathematics of Data Science 3, 854-872 (2021). Supplemental material here.
• Markov chain Monte Carlo algorithms with sequential proposals. Joint with Joonha Park. Statistics and Computing 30, 1325–1345 (2020).
• Automatic adaptation of MCMC algorithms. (File in pdf). Joint with Dao Nguyen, Perry de Valpine, Daniel Turek, Nicholas Michaud, Christopher Paciorek. Bayesian Analysis 15 (4) 1323 - 1343 (2020).
• Efficiency bounds for semiparametric models with singular score functions. (File in pdf). Jointly with P. Dovonon. Econometric Review 39, 612-648 (2020).
• Unbiased Markov Chain Monte Carlo with couplings. Joint with Pierre Jacob, and John O'Leary. JRSS-B with discussion (2020).
• Sequential change-point detection in highdimensional Gaussian graphical models. Joint work with Hossein Keshararz and George Michailidis. JMLR 21 (82), 1-57 (2020).
• Likelihood Inference for Large Scale Stochastic Blockmodels with Covariates based on a Divide-and-Conquer Parallelizable Algorithm with Communication. (File in pdf). JCGS 3, vol 28 (2019). Joint with Sandipan Roy, and George Michailidis.
• A quasi-Bayesian estimation of large Gaussian graphical models. (File in pdf). Journal of Multivariate Analysis 173, 656-671 (2019). A Matlab implementation is available here .
•  Stochastic FISTA algorithms : so fast ? (File in pdf). Joint with Gersende Fort,  Laurent Risser, and Eric Moulines. Proceeding of the 2018 IEEE Statistical Signal Processing Workshop (2018).
• A scalable algorithm for Gaussian graphical models with change-points, joint with Leland Bybee.  JMLR 19, 440-477 (2018).
• On the contraction properties of some high-dimensional quasi-posterior distributions. (File in pdf). Annals of Statistics 45, 2248-2273 (2017).
  
• On stochastic Proximal Gradient Algorithms, joint with Eric Moulines, and Gersende Fort. JMLR 18, 1-33 (2017).
• Change-point Estimation in High-dimensional Markov Random Fields, with a supplement. Joint with Sandipan Roy, and George Michailidis. JRSS-B 79, 1187-1206 (2017).
  
• On the sparse Bayesian learning of linear models. With Chai Chye Yee. Communications in Statistics – Theory and Methods 46, 7672-7691 (2017).
  
• Markov Chain Monte Carlo Confidence Intervals. (File in pdf). Bernoulli 22 (3), 1808-1838, (2016).
• On Russian Roulette Estimates for Bayesian inference with doubly-intractable Likelihoods, joint with M. Girolami, A-M. Lyne, H. Strathmann and D. Simpson. A very nice web-illustration of the method is here. Statistical Science, Vol. 30, No. 4, 443-467 (2015).
  
• Inference for dynamic and latent variable models via iterated, perturbed Bayes maps. With E. Ionides, D. Nguyen, S. Stoev, and A. King. PNAS 112 (3) 719-724 (2015).
  
• On the Convergence Rates of Some Adaptive Markov Chain Monte Carlo Algorithms. (file in pdf ) Joint work work with Yizao Wang.  Journal of Applied Probability, 52, 811-825 (2015).
  
• Estimation of Network structures from partially observed Markov random fields. (File in pdf). Electronic Journal of Statistics, Vol 8, Number 2, 2242-2263 (2014).
  
• A martingale decomposition for quadratic forms of Markov chains (with Applications). With Matias Cattaneo. (File in pdf). Stochastic Processes and their Applications, Vol. 124, Issue 1 646-677, (2014).
  
• Discussion of A Brief Survey of Modern Optimization for Statisticians (jointly with George Michailidis) a discussion of the paper "A Brief Survey of Modern Optimization for Statisticians", by K. Lange, E. C. Chi, and H. Zhou, to appear in International Statistical Review, 82, 1, 71-75 (2014).
• Bayesian inference of exponential random graph models for large social networks. With Jing Wang. Communications in Statistics - Simulation and Computation,  Vol. 43, 359-377 (2014). Here is a pdf version.
• Bayesian computation for statistical models with intractable normalizing constants. With Nicolas Lartillot and Christian Robert (here is a version in pdf). Brazilian Journal of Probability and Statistics, Vol. 27, 416-436 (2013).
• Modeling choice interdependence in a social network. With Jing Wang and Anocha Aribarg. Marketing Science, Vol. 32 Issue 6, 977-997 (2013).
• Limit Theorems for some adaptive MCMC algorithms with sub-geometric kernels: Part II. With Gersende Fort. (here is a pdf version). Bernoulli Vol. 18, 975-1001 (2012).
• Iterated Filtering. With Edward Ionides, Anindya Bhadra and Aaron King. Annals of Statistics, Vol. 39, 1776-1802 (2011).
• Optimal scaling of Metropolis-Coupled Markov chain Monte Carlo. With Jeff Rosenthal and Gareth Roberts. (file in pdf). Statistics and Computing 21, No 4, page 555-568 (2011).
• A computational framework for empirical Bayes inference. (file in pdf). Statistics and Computing 21, No 4, page 463-473 (2011).
• Kernel Estimators of asymptotic variance for adaptive Markov Chain Monte Carlo (file in pdf, with a supplement paper). Annals of Statistics 39, No 2, 990-1011 (2011).
• Adaptive Markov Chain Monte Carlo: Theory and Methods. Joint with Gersende Fort, Eric Moulines and Pierre Priouret. (file in pdf). Book Chapter in "Bayesian Time Series Models". Cambridge University Press, 2011.
• A cautionary tale on the efficiency of some adaptive Monte Carlo schemes. (file in pdf). Annals of Applied Probability 20, No 3, 116-154, 2010.
• Limit Theorems for some adaptive MCMC algorithms with sub-geometric kernels. With Gersende Fort. (file in pdf). Bernoulli 16 (2010), 116-154.
• The Wang-Landau algorithm in general state spaces: applications and convergence analysis. with Jun Liu. (file in pdf). Statistica Sinica 20 (2010), 209-233.
• Resampling from the past to improve on MCMC algorithms (file in pdf). Far East Journal of Theoretical Probability 27, Issue 1 (2009).
• On the efficiency of adaptive MCMC algorithms, with Christophe Andrieu. Elect. Comm. In Prob. 12 (2007), 336-349. (file in pdf).
• On the geometric ergodicity of Metropolis-Hastings algorithms, with Francois Perron, Statistics 41 (2007), 77-84. (file in pdf).
• Discussion of the Equi-Energy sampler, with Jun Liu, Annals of Statistics 34 (2006), No. 4. (file in pdf).
• An adaptive version for the Metropolis adjusted Langevin algorithm with a truncated drift. Methodology and Computing in Applied Probability 8 (2006), 235-254.  (file in pdf)  Supplementary file for one of the simulation examples.
• On adaptive Markov chain Monte Carlo algorithms, with Jeff Rosenthal, Bernoulli 11 (2005), 815-828. (file in pdf).
• Improving on the Independent Metropolis algorithm, with Francois Perron, Statistica Sinica 15 (2005), 3-18. (file in pdf).

Some unpublished manuscripts

• Hyperspectral image unmixing: accounting for wavelength dependence. (File in pdf). Joint work with  Chia Chye Yee.
  
• Randomized Evaluation of Institutions: Theory with Application to Voting and Deliberation Experiments. With Leonard Wantchekon (file in pdf).
• Ethnic Solidarity and National Public Goods Programmes: Evidence from a Field Experiment with Gwyneth McClendon, and Leonard Wantchekon (here is a pdf version ).
• A strong law of large numbers for martingale arrays. (file in pdf).
• Scalable Computation of Regularized Precision Matrices via Stochastic Optimization. (File in pdf). (2016) Joint with Rahul Mazumder, and Jie Chen.
  
• A Moreau-Yosida approximation for high-dimensional posterior  distributions (2017). (File in pdf).   A Matlab implementation of some of the methods is available here.