Papers etc

Lecture notes

• • Bayesian Nonparametrics [PDF]
  • Machine Learning (Slides) [PDF]
  • Probability Theory II [PDF]

Papers

• Spectral Representations for Accurate Causal Uncertainty Quantification with Gaussian Processes.
  H Dance, P Orbanz and A Gretton.
  [arxiv]
• Slow rates of approximation of U-statistics and V-statistics by quadratic forms of Gaussians.
  KH Huang and P Orbanz.
  [arxiv]
• Designing Mechanical Meta-Materials by Learning Equivariant Flows.
  M Mirramezani, AS Meeussen, K Bertoldi, P Orbanz, and RP Adams.
  [arxiv]
• Gaussian universality for approximately polynomial functions of high-dimensional data.
  KH Huang, M Austern, and P Orbanz.
  [arxiv]
• Gaussian and non-Gaussian universality of data augmentation.
  KH Huang, P Orbanz, and M Austern.
  [arxiv]
• Global optimality under amenable symmetry constraints.
  P Orbanz.
  [arxiv]
• Poisson approximation for stochastic processes summed over amenable groups.
  H Ye, P Orbanz, and M Austern.
  [arxiv]
• Representing and learning functions invariant under crystallographic groups.
  With RP Adams.
  [arxiv]
• The Graphical Matrix Pencil Method: Exchangeable Distributions with Prescribed Subgraph Densities.
  LM Gunderson, G Bravo-Hermsdorff, and P Orbanz.
  NeurIPS 2023.
  [Paper]
• Limit theorems for distributions invariant under groups of transformations.
  M Austern and P Orbanz.
  Annals of Statistics, Vol. 50, No. 4, 1960–1991, 2022.
  [PDF] [Journal]
• Uniform estimation in stochastic block models is slow.
  With I Castillo.
  Electronic Journal of Statistics, Vol. 16, No. 1, 2947-3000, 2022.
  [arxiv] [Journal]
• Empirical risk minimization and stochastic gradient descent for relational data.
  V Veitch, M Austern, W Zhou, DM Blei and P Orbanz.
  AISTATS 2019.
  [arxiv] [Code]
• Distribution theory for hierarchical processes.
  With F Camerlenghi, A Lijoi and I Prünster.
  Annals of Statistics, Vol. 47, 67-92, 2019.
  [Journal]
• Non-vacuous Generalization Bounds at the ImageNet Scale: a PAC-Bayesian Compression Approach.
  W Zhou, V Veitch, M Austern, RP Adams and P Orbanz.
  ICLR 2019.
  [arxiv]
• Random walk models of network formation and sequential Monte Carlo methods for graphs.
  B Bloem-Reddy and P Orbanz.
  Journal of the Royal Statistical Society, Series B, 80, 871-898, 2018.
  [Journal] [arxiv] [Code] [A talk on this work]
• Discussion of F. Caron and E. B. Fox, "Sparse graphs using exchangeable random measures."
  P Orbanz.
  [Discussion] Journal of the Royal Statistical Society, Series B, 79, Part 5.
  [PDF]
• Subsampling large graphs and invariance in networks.
  P Orbanz.
  [arxiv] [A talk on this work]
• Preferential attachment and vertex arrival times.
  With B Bloem-Reddy.
  [arxiv] [An article on inference in these models]
• Independence by random scaling.
  With LF James.
  [arxiv]
• Borel liftings of graph limits.
  With B Szegedy.
  Electronic Communications in Probability, Vol. 21, paper no. 65, 2016.
  [PDF]
• Bayesian models of graphs, arrays and other exchangeable random structures.
  With DM Roy.
  IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 37, 437-461, 2015.
  [Journal] [arxiv]
  Some of the key ideas described in this paper are also explained in the following talk:
  [Video]
• Scaled subordinators and generalizations of the Indian buffet process.
  With LF James and YW Teh.
  [arxiv]
• Nonparametric priors on complete separable metric spaces.
  P Orbanz.
  [PDF]
• Random function priors for exchangeable arrays with applications to graphs and relational data.
  JR Lloyd, P Orbanz, Z Ghahramani and DM Roy.
  Advances in Neural Information Processing Systems, 2012.
  [PDF]   [Talk on this and related work (Video, 50min)]
• Projective Limit Random Probabilities on Polish Spaces.
  P Orbanz.
  Electronic Journal of Statistics, Vol. 5, 1354-1373, 2011.
  [PDF]
• Unit-rate Poisson representations of completely random measures.
  With S Williamson.
  [PDF]
• Dependent Indian Buffet Processes.
  S Williamson, P Orbanz and Z Ghahramani.
  AISTATS 2010, JMRL W&CP 9:924-931.
  [PDF]
• Conjugate Projective Limits.
  P Orbanz.
  [arxiv]
• Bayesian Nonparametric Models.
  P Orbanz and YW Teh.
  In Encyclopedia of Machine Learning. Springer, 2010.
  [PDF]
• Construction of Nonparametric Bayesian Models from Parametric Bayes Equations.
  P Orbanz.
  Advances in Neural Information Processing Systems, 2009.
  [PDF]   [Supplements (Proofs)]
  [Techreport Version] (Identical text; proofs included as appendix)
• Functional Conjugacy in Parametric Bayesian Models.
  P Orbanz.
  Techreport, 2009.
  [PDF]
• Music Preference Learning with Partial Information.
  Y Moh, P Orbanz and JM Buhmann.
  ICASSP 2008.
  [PDF]
• Nonparametric Bayesian Image Segmentation.
  P Orbanz and JM Buhmann.
  International Journal of Computer Vision (IJCV), Vol. 77, 25-45, 2008.
  [PDF]   [Journal]
• Cluster Analysis of Heterogeneuos Rank Data.
  LM Busse, P Orbanz and JM Buhmann.
  International Conference on Machine Learning (ICML), 2007.
  [PDF]   [PDF (with corrections)]
• Bayesian Order-Adaptive Clustering for Video Segmentation.
  P Orbanz, S Braendle and JM Buhmann.
  EMMCVPR, 2007.
  [PDF]   [Publisher]
• Smooth Image Segmentation by Nonparametric Bayesian Inference.
  P Orbanz and JM Buhmann.
  European Conference on Computer Vision (ECCV), Vol. 1, 444-457, 2006.
  [PDF]   [Publisher]
• SAR Images as Mixtures of Gaussian Mixtures.
  P Orbanz and JM Buhmann.
  IEEE International Conference on Image Processing (ICIP), Vol. 2, 209-212, 2005.
  [PDF]   [Publisher]