Coming Soon Refereed Theses Notes Translations

Coming Soon

  • Nguyen, TD, Giordano, R, Meager, R, and Broderick, T.
    Sensitivity of MCMC-based analyses to small-data removal. Preprint on arXiv:2408.07240 [stat.ME] [arXiv]
  • Huang, JY, Burt, DR, Nguyen, TD, Shen, Y, and Broderick, T.
    Approximations to worst-case data dropping: unmasking failure modes. Preprint on arXiv:2408.09008 [stat.ME] [arXiv]
  • Shen, Y*, Berlinghieri, R*, and Broderick, T. (*equal contribution)
    Multi-marginal Schrödinger Bridges with Iterative Reference Refinement. Preprint on arXiv:2408.06277 [stat.ML] [arXiv]
  • Shen, Y, Masoero, L, Schraiber, J, and Broderick, T.
    Double trouble: Predicting new variant counts across two heterogeneous populations. Preprint on arXiv:2403.02154 [stat.ME] [arXiv]
  • Burt, DR, Shen, Y, and Broderick, T.
    Consistent Validation for Predictive Methods in Spatial Settings. Preprint on arXiv:2402.03527 [stat.ML] [arXiv]
  • Shiffman, M, Giordano, R, and Broderick, T.
    Could dropping a few cells change the takeaways from differential expression? Preprint on arXiv:2312.06159 [q-bio.QM] [arXiv]
  • Kasprzak, MJ, Giordano, R, and Broderick, T.
    How good is your Laplace approximation of the Bayesian posterior? Finite-sample computable error bounds for a variety of useful divergences. Preprint on arXiv:2209.14992 [math.ST] [arXiv] [code]
  • Giordano, R and Broderick, T.
    The Bayesian Infinitesimal Jackknife for Variance. Preprint on arXiv:2305.06466 [stat.ME] [arXiv]
  • Broderick, T, Giordano, R, and Meager, R. (alphabetical authorship)
    An Automatic Finite-Sample Robustness Metric: Can Dropping a Little Data Change Conclusions? Preprint on arXiv:2011.14999 [stat.ME] [arXiv] [paper code] [method code] [youtube]
  • Giordano, R, Jordan, MI, and Broderick, T.
    A Higher-Order Swiss Army Infinitesimal Jackknife. Preprint on arXiv:1907.12116 [math.ST] [arXiv]

Refereed

  • Giordano, R*, Ingram, M*, and Broderick, T. (*equal contribution)
    Black Box Variational Inference with a Deterministic Objective: Faster, More Accurate, and Even More Black Box. Journal of Machine Learning Research, 2024. [link] [arXiv] [code] [MIT News]
  • Deshpande, SK*, Ghosh, S*, Nguyen, TD*, and Broderick, T. (*equal contribution)
    Are you using test log-likelihood correctly? Transactions on Machine Learning Research, 2024. [TMLR Infinite Conference: pdf, openreview, video] [arXiv] [code] [youtube]
  • Bonaker, N, Nel, E, Vertanen, K, and Broderick, T.
    A Usability Study of Nomon: A Flexible Interface for Single-Switch Users. ACM SIGACCESS Conference on Computers and Accessibility (ASSETS), 2023. [link] [demo, code, Nomon website]
  • Berlinghieri, R, Trippe, BL, Burt, DR, Giordano, R, Srinivasan, K, Özgökmen, T, Xia, J, and Broderick, T.
    Gaussian processes at the Helm(holtz): A more fluid model for ocean currents. ICML, 2023. [link] [arXiv] [code] [MIT News]
  • Nguyen, T, Huggins, J, Masoero, L, Mackey, L, and Broderick, T.
    Independent finite approximations for Bayesian nonparametric inference [Previously "Independent finite approximations for Bayesian nonparametric inference: construction, error bounds, and practical implications"]. Bayesian Analysis, 2023. [link] [arXiv] [youtube]
  • Campbell, T, Syed, S, Yang, C, Jordan, MI, and Broderick, T.
    Local Exchangeability. Bernoulli, 2023. [link] [arXiv]
  • Trippe, BL, Yim, J, Tischer, D, Broderick, T., Baker, D, Barzilay, R, and Jaakkola, T.
    Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem. ICLR, 2023. [link] [arXiv] [code]
  • Giordano, R*, Liu, R*, Jordan, MI, and Broderick, T. (*equal contribution)
    Evaluating Sensitivity to the Stick-Breaking Prior in Bayesian Nonparametrics (with Discussion). Bayesian Analysis, 2023. [link] [arXiv main paper] [arXiv rejoinder] [code] [youtube]
    Earlier version:
    • Liu, R*, Giordano, R*, Jordan, MI, and Broderick, T. (*equal contribution)
      Evaluating sensitivity to the stick breaking prior in Bayesian nonparametrics. BNP@NeurIPS. 2018. [link pdf] [arXiv] [workshop link]
  • Trippe, BL, Deshpande, SK, and Broderick, T.
    Confidently comparing estimators with the c-value. Journal of the American Statistical Association, 2023. [link] [arXiv] [code] [MIT News]
  • Broderick, T, Gelman, A, Meager, R, Smith, AL, and Zheng, T.
    Toward a Taxonomy of Trust for Probabilistic Machine Learning. Science Advances, 2023. [link] [arXiv] [MIT News]
  • Agrawal, R and Broderick, T.
    The SKIM-FA Kernel: High-Dimensional Variable Selection and Nonlinear Interaction Discovery in Linear Time. Journal of Machine Learning Research, 2023. [link] [arXiv] [code]
  • Camerlenghi, F, Favaro, S, Masoero, L, and Broderick, T.
    Scaled process priors for Bayesian nonparametric estimation of the unseen genetic variation. Journal of the American Statistical Association, 2022. [link] [arXiv] [code]
  • Bonaker, N, Nel, E, Vertanen, K, and Broderick, T.
    A Performance Evaluation of Nomon: A Flexible Interface for Noisy Single-Switch Users. ACM CHI Conference on Human Factors in Computing, 2022. [link] [arXiv] [demo, code, Nomon website] [youtube] [MIT News]
  • Bonaker, N, Nel, E, Vertanen, K, and Broderick, T.
    Demonstrating Nomon: A Flexible Interface for Noisy Single-Switch Users. CHI Interactivity 2022. [link] [demo, code, Nomon website] [youtube]
  • Nguyen, TD, Trippe, BL, and Broderick, T.
    Many processors, little time: MCMC for partitions via optimal transport couplings. AISTATS 2022. [link] [arXiv] [code] [video and slides at link]
    Earlier version:
    • Trippe, BL*, Nguyen, TD*, and Broderick, T. (*equal contribution)
      Optimal Transport Couplings of Gibbs Samplers on Partitions for Unbiased Estimation. 3rd Symposium on Advances in Approximate Bayesian Inference. 2021. [workshop link] [link] [arXiv] [code] [youtube]
  • Stephenson, WT, Ghosh, S, Nguyen, TD, Yurochkin, M, Deshpande, SK, and Broderick, T.
    Measuring the robustness of Gaussian processes to kernel choice. AISTATS 2022. [link] [arXiv] [code]
  • Stephenson, WT, Frangella, Z, Udell, M, and Broderick, T.
    Can we globally optimize cross-validation loss? Quasiconvexity in ridge regression. Neural Information Processing Systems. 2021. [link] [arXiv] [code zip] [youtube]
  • Trippe, BL, Finucane, HK, and Broderick, T.
    For high-dimensional hierarchical models, consider exchangeability of effects across covariates instead of across datasets. Neural Information Processing Systems. 2021. [link] [arXiv] [youtube]
  • Cai, D*, Campbell, T*, and Broderick, T. (*equal contribution)
    Finite mixture models do not reliably learn the number of components. [Previously "Finite mixture models are typically inconsistent for the number of components"]. ICML. 2021. [link] [arXiv] [video at link]
  • Agrawal, R and Broderick, T.
    High-Dimensional Variable Selection and Non-Linear Interaction Discovery in Linear Time. SubSetML: Subset Selection in Machine Learning: From Theory to Practice. ICML Workshop, 2021. [workshop link] [arXiv]
  • Giordano, R, Meager, R, and Broderick, T.
    An Automatic Finite-Sample Robustness Metric: Can Dropping a Little Data Change Conclusions? Workshop on Distribution-Free Uncertainty Quantification at ICML, 2021. [workshop link] [arxiv] [code] [youtube]
  • Masoero, L, Camerlenghi, F, Favaro, S, and Broderick, T.
    More for less: Predicting and maximizing genetic variant discovery via Bayesian nonparametrics. Biometrika. 2021. [link] [arXiv] [code] [youtube]
    Earlier version:
    • Masoero, L, Camerlenghi, F, Favaro, S, and Broderick, T.
      Genomic variety prediction via Bayesian nonparametrics. Symposium on Advances in Approximate Bayesian Inference. 2019. [workshop link]
  • Cai, D*, Campbell, T*, and Broderick, T. (*equal contribution)
    Power posteriors do not reliably learn the number of components in a finite mixture. I Can't Believe It's Not Better! NeurIPS Workshop. 2020. [workshop link]
  • Nguyen, TD, Huggins, JH, Masoero, L, Mackey, L, and Broderick, T.
    Independent versus truncated finite approximations for Bayesian nonparametric inference. I Can't Believe It's Not Better! NeurIPS Workshop. 2020. [workshop link]
  • Stephenson, WT, Udell, M, and Broderick, T.
    Approximate Cross-Validation with Low-Rank Data in High Dimensions. Neural Information Processing Systems. 2020. [link] [arXiv] [code]
  • Ghosh, S*, Stephenson, WT*, Nguyen, TD, Deshpande, SK, and Broderick, T. (*equal contribution)
    Approximate Cross-Validation for Structured Models. Neural Information Processing Systems. 2020. [link] [arXiv] [code] [youtube]
  • Haibe-Kains, B, Adam, GA, Hosny, A, Khodakarami, F, Massive Analysis Quality Control (MAQC) Society Board of Directors, Waldron, L, Wang, B, McIntosh, C, Goldenberg, A, Kundaje, A, Greene, CS, Broderick, T, Hoffman, MM, Leek, JT, Korthauer, K, Huber, W, Brazma, A, Pineau, J, Tibshirani, R, Hastie, T, Ioannidis, JPA, Quackenbush, G, Aerts, HJWL
    Transparency and reproducibility in artificial intelligence. Nature, 2020. [link]
  • Stephenson, W and Broderick, T.
    Approximate Cross-Validation in High Dimensions with Guarantees [Previously "Sparse Approximate Cross-Validation for High-Dimensional GLMs"]. AISTATS, 2020. [link] [arXiv] [code] [youtube]
    Earlier version:
    • Stephenson, W and Broderick, T.
      Approximate Cross-Validation in High Dimensions with Guarantees. NeurIPS 2019 Workshop on Machine Learning with Guarantees. 2019. [workshop link]
  • Huggins, JH, Kasprzak, M, Campbell, T, and Broderick, T.
    Validated Variational Inference via Practical Posterior Error Bounds. AISTATS, 2020. [link] [arXiv] [code]
    Some related early work appears in:
    • Huggins, JH, Campbell, T, Kasprzak, M, and Broderick, T.
      Practical bounds on the error of Bayesian posterior approximations: A nonasymptotic approach. Preprint on arXiv:1809.09505 [math.ST]. 2018. [arXiv]
  • Agrawal, R, Huggins, JH, Trippe, B, and Broderick, T.
    The Kernel interaction trick: Fast Bayesian discovery of pairwise interactions in high dimensions. ICML. 2019. [link] [arXiv] [original code] [Pyro code] [youtube]
  • Trippe, B, Huggins, JH, Agrawal, R, and Broderick, T.
    LR-GLM: High-dimensional Bayesian inference using low-rank data approximations. ICML. 2019. [link] [arXiv] [code] [slideslive video: starts at 31:41]
    Earlier version:
    • Trippe, B, Huggins, JH, and Broderick, T.
      Fast Bayesian inference in GLMs with low rank data approximations. Symposium on Advances in Approximate Bayesian Inference. 2018. [link pdf] [workshop link]
  • Giordano, R, Stephenson, W, Liu, R, Jordan, MI, and Broderick, T.
    A Swiss army infinitesimal jackknife [Previously "Return of the infinitesimal jackknife"]. AISTATS. 2019. (Notable Paper Award) [link] [arXiv] [code] [youtube]
  • Agrawal, R, Campbell, T, Huggins, JH, and Broderick, T.
    Data-dependent compression of random features for large-scale kernel approximation. AISTATS. 2019. [link] [arXiv] [code]
  • Huggins, JH, Campbell, T, Kasprzak, M, and Broderick, T.
    Scalable Gaussian process inference with finite-data mean and variance guarantees. AISTATS. 2019. [link] [arXiv]
  • Campbell, T*, Huggins, JH*, How, J, and Broderick, T. (*equal contribution)
    Truncated random measures. Bernoulli. 2019. [link] [arXiv]
  • Campbell, T and Broderick, T.
    Automated scalable Bayesian inference via Hilbert coresets. Journal of Machine Learning Research. 2019. [link] [arXiv] [code] [youtube]
  • Giordano, R, Broderick, T, and Jordan, MI.
    Covariances, robustness, and variational Bayes. Journal of Machine Learning Research. 2018. [link] [arXiv] [code] [youtube]
    Earlier versions:
    • Giordano, R, Broderick, T, and Jordan, MI.
      Fast Measurements of Robustness to Changing Priors in Variational Bayes. NeurIPS 2016 Workshop on Advances in Approximate Bayesian Inference. 2016. [link pdf] [workshop link] [arXiv]
    • Giordano, R, Broderick, T, and Jordan, MI.
      Robust inference with variational Bayes. NeurIPS 2015 Workshop on Advances in Approximate Bayesian Inference. 2015. [link pdf] [workshop link] [arXiv]
    • Giordano, R and Broderick, T.
      Covariance matrices and influence scores for mean field variational Bayes. Preprint on arXiv:1502.07685 [stat.ML] [arXiv]
  • Masoero, L*, Stephenson, W*, and Broderick, T.
    Sensitivity of Bayesian inference to data perturbations. Symposium on Advances in Approximate Bayesian Inference. 2018. [link pdf] [workshop link]
  • Masoero, L, Camerlenghi, F, Favaro, S, and Broderick, T.
    Posterior representations of hierarchical completely random measures in trait allocation models. BNP@NeurIPS. 2018. [link pdf] [workshop link]
  • Campbell, T, Cai, D, and Broderick, T.
    Exchangeable trait allocations. Electronic Journal of Statistics. 2018. [link] [arXiv]
  • Agrawal, R, Broderick, T, and Uhler, C.
    Minimal I-MAP MCMC for scalable structure discovery in causal DAG models. ICML. 2018. [link] [arXiv] [code]
  • Campbell, T and Broderick, T.
    Bayesian coreset construction via Greedy Iterative Geodesic Ascent. ICML. 2018. [link] [arXiv] [code] [youtube]
  • Broderick, T, Wilson, AC, and Jordan, MI.
    Posteriors, conjugacy, and exponential families for completely random measures. Bernoulli. 2018. [link] [arXiv]
  • Shiffman, M, Stephenson, W, Schiebinger, G, Campbell, T, Huggins, JH, Regev, A, and Broderick, T.
    Probabilistic reconstruction of cellular differentiation trees from single-cell RNA-seq data. NeurIPS Workshop on Machine Learning in Computational Biology. 2017. [workshop link]
  • Cai, D, Campbell, T, and Broderick, T.
    Finite mixture models are typically inconsistent for the number of components. NeurIPS 2017 Workshop on Advances in Approximate Bayesian Inference. 2017. [link pdf] [workshop link]
  • Giordano, R*, Liu, R*, Varoquaux, N*, Jordan, MI, and Broderick, T. Measuring cluster stability for Bayesian non-parametrics using the linear bootstrap. NeurIPS 2017 Workshop on Advances in Approximate Bayesian Inference. 2017. [link pdf] [workshop link]
  • Huggins, JH, Masoero, L, Mackey, L, and Broderick, T.
    Generic finite approximations for practical Bayesian nonparametrics. NeurIPS 2017 Workshop on Advances in Approximate Bayesian Inference. 2017. [link pdf] [workshop link]
  • Shiffman, M, Stephenson, W, Schiebinger, G, Campbell, T, Huggins, JH, Regev, A, and Broderick, T.
    Probabilistic reconstruction of cellular differentiation trees from single-cell RNA-seq data. NeurIPS 2017 Workshop on Advances in Approximate Bayesian Inference. 2017. [poster pdf] [workshop link]
  • Huggins, JH, Adams, RP, and Broderick, T.
    PASS-GLM: polynomial approximate sufficient statistics for scalable Bayesian GLM inference. Neural Information Processing Systems. 2017. [link] [arXiv]
  • Stephenson, W and Broderick, T.
    Understanding Covariance Estimates in Expectation Propagation. NeurIPS 2016 Workshop on Advances in Approximate Bayesian Inference. 2016. [link pdf] [workshop link]
  • Guo, F, Wang, X, Fan, K, Broderick, T, and Dunson, D.
    Boosting Variational Inference. NeurIPS 2016 Workshop on Advances in Approximate Bayesian Inference. 2016. [link pdf] [workshop link] [arXiv]
  • Cai, D, Campbell, T, and Broderick, T.
    Paintboxes and probability functions for edge-exchangeable graphs. NeurIPS 2016 Workshop on Adaptive and Scalable Nonparametric Methods in Machine Learning. 2016. [link pdf] [workshop link]
  • Campbell, T, Cai, D, and Broderick, T.
    A Paintbox Representation of Exchangeable Trait Allocations. NeurIPS 2016 Workshop on Practical Bayesian Nonparametrics. 2016. [link pdf] [workshop link]
  • Huggins, JH, Campbell, T, and Broderick, T.
    Coresets for scalable Bayesian logistic regression. Neural Information Processing Systems. 2016. [link] [arXiv] [spotlight video and more info]
  • Cai, D, Campbell, T, and Broderick, T.
    Edge-exchangeable graphs and sparsity. Neural Information Processing Systems. 2016. [link] [spotlight video and more info]
    Earlier versions:
    • Cai, D, Broderick, T.
      Completely random measures for modeling power laws in sparse graphs. NeurIPS 2015 Workshop on Networks in the Social and Information Sciences. 2015. [link pdf] [workshop link] [arXiv]
    • Broderick, T, and Cai, D.
      Edge-exchangeable graphs and sparsity. NeurIPS 2015 Workshop on Networks in the Social and Information Sciences. 2015. [link pdf] [workshop link] [arXiv]
    • Broderick, T, and Cai, D.
      Edge-exchangeable graphs, sparsity, and power laws. NeurIPS 2015 Workshop: Bayesian Nonparametrics: The Next Generation. 2015. [link pdf] [workshop link]
  • Giordano, R, Broderick, T, Meager, R, Huggins, JH, and Jordan, MI.
    Fast robustness quantification with variational Bayes. ICML 2016 Workshop on #Data4Good: Machine Learning in Social Good Applications. 2016. [workshop link] [arXiv]
  • Giordano, R, Broderick, T, and Jordan, MI.
    Linear response methods for accurate covariance estimates from mean field variational Bayes. Neural Information Processing Systems. 2015. [link] [workshop link] [arXiv] [code] [youtube]
    Earlier versions:
    • Giordano, R and Broderick, T.
      Covariance matrices and influence scores for mean field variational Bayes. Preprint on arXiv:1502.07685 [stat.ML] [arXiv]
    • Giordano, R and Broderick, T.
      Covariance matrices for mean field variational Bayes. NeurIPS 2014 Workshop on Advances in Variational Inference. 2014. [link pdf] [workshop link] [arXiv]
  • Broderick, T, Mackey, L, Paisley, J, and Jordan, MI.
    Combinatorial clustering and the beta negative binomial process. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2015 (2011 arXiv). [link] [arXiv] [code: zip]
  • Broderick, T and Steorts, RC.
    Variational Bayes for merging noisy databases. NeurIPS 2014 Workshop on Advances in Variational Inference. 2014. [link] [workshop link] [arXiv]
  • Luts, J, Broderick, T, and Wand, MP.
    Real-time semiparametric regression. Journal of Computational and Graphical Statistics. 2014. [link] [arXiv]
  • Broderick, T, Pitman, J, and Jordan, MI.
    Feature allocations, probability functions, and paintboxes. Bayesian Analysis. 2013. [linked pdf] [arXiv]
  • Broderick, T, Boyd, N, Wibisono, A, Wilson, AC, and Jordan, MI.
    Streaming variational Bayes. Neural Information Processing Systems. 2013. [link] [arXiv] [code]
  • Pan, X, Gonzalez, JE, Jegelka, S, Broderick, T, and Jordan, MI.
    Optimistic concurrency control for distributed unsupervised learning. Neural Information Processing Systems. 2013. [link] [arXiv]
  • Broderick, T, Jordan, MI, and Pitman, J.
    Cluster and feature modeling from combinatorial stochastic processes. Statistical Science. 2013. [link] [arXiv]
  • Broderick, T, Kulis, B, and Jordan, MI.
    MAD-Bayes: MAP-based asymptotic derivations from Bayes. ICML. 2013. [link] [arXiv] [code]
  • Broderick, T, Jordan, MI, and Pitman, J.
    Beta processes, stick-breaking, and power laws. Bayesian Analysis. 2012. [link] [arXiv]
  • Morgan, AN, Long, J, Richards, JW, Broderick, T, Butler, NR, Bloom, JS.
    Rapid, machine-learned resource allocation: application to high-redshift GRB follow-up. Astrophysical Journal. 2012. [link] [arXiv]
  • Broderick, T and Gramacy, RB.
    Classification and categorical inputs with treed Gaussian process models. Journal of Classification. 2011. [link] [arXiv]
  • Broderick, T, Wong-Lin, KF, and Holmes, P.
    Closed-form approximations of first-passage distributions for a stochastic decision making model. Applied Mathematics Research Express. 2010. [link]
  • Kapicioglu, B, Schapire, RE, Wikelski, M, and Broderick, T.
    Combining spatial and telemetric features for learning animal movement models. Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence. 2010. [linked pdf]
  • Broderick, T and MacKay, DJC.
    Fast and flexible selection with a single switch. PLoS ONE 4(10), e7481. 2009. [link] [arXiv] [further resources and code link]
  • Broderick, T and Gramacy, RB.
    Treed Gaussian process models for classification. Proceedings of the 11th IFCS Biennial Conference. 2009. [Springer]
  • Mandelbaum, R, Hirata, CM, Broderick, T, Seljak, U, and Brinkmann, J.
    Ellipticity of dark matter halos with galaxy-galaxy weak lensing. Monthly Notices of the Royal Astronomical Society 370, 1008-1024. 2006. [arXiv]
  • Huterer, D, Kim, A, Krauss, LM, and Broderick, T.
    Redshift accuracy requirements for future supernova and number count surveys. Astrophysical Journal 615, 595-602. 2004. [arXiv]

Theses

  • Broderick, T.
    Clusters and features from combinatorial stochastic processes. PhD Dissertation. Department of Statistics, UC Berkeley. 2014.
  • Broderick, T.
    Nomon: Efficient communication with a single switch. Technical Report (extension to Master's Thesis). Cavendish Laboratory, University of Cambridge. 2009. [pdf]
  • Broderick, T.
    Treed models and Gaussian processes for classification. Master of Advanced Study in Mathematics (Part III) Essay. Department of Pure Mathematics and Mathematical Statistics, University of Cambridge. 2008.
  • Broderick, T.
    Construction of a pairwise Ising distribution over a large state space with sparse data. Senior Thesis. Mathematics Department, Program in Applications of Computing, Program in Applied and Computational Mathematics, Princeton University. 2007.

Technical Reports and Notes

  • Shiffman, M, Stephenson, WT, Schiebinger, G, Huggins, J, Campbell, T, Regev, A, and Broderick, T.
    Reconstructing probabilistic trees of cellular differentiation from single-cell RNA-seq data. Preprint on arXiv:1811.11790 [q-bio.QM] [arXiv]
  • MacKay, DJC and Broderick, T.
    Probabilities over trees: generalizations of the Dirichlet Diffusion Tree and the Kingman Coalescent. 2007. [link]
  • Broderick, T, Dudik, M, Tkacik, G, Schapire, RE, and Bialek, W.
    Faster solutions of the inverse pairwise Ising problem. Preprint on arXiv:0712.2437 [q-bio.QM]. 2007. [arXiv]

Translations

  • Alvarez-Melis, D and Broderick, T.
    A translation of "The characteristic function of a random phenomenon" by Bruno de Finetti. On arXiv:1512.01229 [math.ST] [arXiv]
Accessibility
Plain Academic