NeurIPS workshop on Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems

New Orleans, Louisiana, United States

December 2, 2022

Abstract

In recent years, the growth of decision-making applications, where principled handling of uncertainty is of key concern, has led to increased interest in Bayesian techniques. By offering the capacity to assess and propagate uncertainty in a principled manner, Gaussian processes have become a key technique in areas such as Bayesian optimization, active learning, and probabilistic modeling of dynamical systems. In parallel, the need for uncertainty-aware modeling of quantities that vary over space and time has led to large-scale deployment of Gaussian processes, particularly in application areas such as epidemiology. In this workshop, we bring together researchers from different communities to share ideas and success stories. By showcasing key applied challenges, along with recent theoretical advances, we hope to foster connections and prompt fruitful discussion. We invite researchers to submit extended abstracts for contributed talks and posters.

Speakers

Schedule

Time (USA)Event
09:00 - 09:30Introduction and Opening Remarks: Zoubin Ghahramani
09:30 - 10:00Invited Talk: Willie Neiswanger
10:00 - 10:30Invited Talk: Marta Blangiardo
10:30 - 11:00Coffee and Discussion
11:00 - 11:30Invited Talk: Viacheslav Borovitskiy
11:30 - 11:45Contributed Talk: Pierre Thodoroff
11:45 - 12:00Contributed Talk: Renato Berlinghieri
12:00 - 13:00Lunch
13:00 - 13:30Lightning Talks
13:30 - 14:00Invited Talk: Jasper Snoek
14:00 - 14:15Contributed Talk: Renzhi Chen
14:15 - 14:45Coffee and Discussion
14:45 - 15:15Invited Talk: Paula Moraga
15:15 - 15:30Contributed Talk: Sulin Liu
15:30 - 15:45Contributed Talk: Andreas Besginow
15:45 - 16:45Poster Session
16:45 - 17:15Invited Talk: Carolina Osorio
17:15 - 17:55Panel Discussion
17:55 - 18:00Closing Remarks

Organizers

Advisory Committee

Sponsors

Accepted Workshop Papers

TitleAuthors
Spatiotemporal modeling of European paleoclimate using doubly sparse Gaussian processes paper posterSeth D Axen (Machine Learning ⇌ Science Colaboratory); Alexandra Gessner (University of Tübingen); Christian Sommer (Heidelberg Academy of Sciences and Humanities); Nils Weitzel (University of Tübingen); Alvaro Tejero-Cantero (University of Tübingen)
Identifying latent climate signals using sparse hierarchical Gaussian processes paperMatt Amos (Lancaster University); Thomas Pinder (Lancaster University); Paul Young (Lancaster University)
Multi-fidelity experimental design for ice-sheet simulation paperPierre Thodoroff (University of Cambridge); Markus Kaiser (University of Cambridge); Rosie Williams (British Antarctic Survey); Robert Arthern (British Antarctic Survey); Scott Hosking (British Antarctic Survey); Neil D Lawrence (University of Cambridge); Ieva Kazlauskaite (University of Cambridge)
c-TPE: Generalizing Tree-structured Parzen Estimator with Inequality Constraints for Continuous and Categorical Hyperparameter Optimization paper posterShuhei Watanabe (University of Freiburg); Frank Hutter (University of Freiburg)
Preferential Bayesian Optimization with Hallucination Believer paper posterShion Takeno (Nagoya Institute of Technology); Masahiro Nomura (CyberAgent, Inc.); Masayuki Karasuyama (Nagoya Institute of Technology)
Symbolic-Model-Based Reinforcement Learning paper posterPierre-Alexandre Kamienny (Facebook AI Research); Sylvain Lamprier (ISIR-SU)
An Active Learning Reliability Method for Systems with Partially Defined Performance Functions paper posterJonathan C Sadeghi (Five AI Ltd); Romain Mueller (Five.ai); John Redford (Five AI Ltd.)
Multi-fidelity Bayesian experimental design using power posteriors paperAndrew Jones (Princeton University Department of Computer Science); Diana Cai (Princeton University); Barbara Engelhardt (Princeton University)
Bayesian Sequential Experimental Design for a Partially Linear Model with a Gaussian Process Prior paperShunsuke Horii (Waseda University)
Fantasizing with Dual GPs in Bayesian Optimization and Active Learning paper posterPaul Chang (Aalto University); Prakhar Verma (Aalto University); ST John (Aalto University & Finnish Center for Artificial Intelligence); Victor Picheny (Prowler); Henry B Moss (Secondmind); Arno Solin (Aalto University)
Deep Gaussian Process-based Multi-fidelity Bayesian Optimization for Simulated Chemical Reactors paper posterTom Savage (Imperial College London); Nausheen Basha (Imperial College London); Omar Matar (Imperial College London); Antonio del Rio Chanona (Imperial College London)
Ice Core Dating using Probabilistic Programming paperAditya Ravuri (University of Cambridge); Tom Andersson (British Antarctic Survey); Ieva Kazlauskaite (University of Cambridge); William Tebbutt (University of Cambridge); Richard E. Turner (University of Cambridge); Scott Hosking (British Antarctic Survey); Neil D Lawrence (University of Cambridge); Markus Kaiser (University of Cambridge)
Scalable Gaussian Process Hyperparameter Optimization via Coverage Regularization paper posterKillian R Wood (University of Colorado, Boulder); Alec  M Dunton (Lawrence Livermore National Labs); Amanda L Muyskens (Lawrence Livermore National Laboratory); Benjamin Priest (Lawrence Livermore National Laboratory)
Integrated Fourier Features for Fast Sparse Variational Gaussian Process Regression paper posterTalay M Cheema (University of Cambridge)
Provably Reliable Large-Scale Sampling from Gaussian Processes paper posterAnthony Stephenson (University of Bristol); Robert Allison (University of Bristol)
Uncovering the short-time dynamics of electricity day-ahead markets paper posterAntonio Malpica Morales (Imperial College London); Serafim Kalliadasis (Imperial College London); Miguel Durán Olivencia (Imperial College London)
Distributionally Robust Bayesian Optimization with φ-divergences paper posterHisham Husain (Amazon); Vu Nguyen (Amazon); Anton van den Hengel (University of Adelaide)
Sparse Bayesian Optimization paper posterSulin Liu (Princeton University); Qing Feng (Facebook); David Eriksson (Meta); Benjamin Letham (Facebook); Eytan Bakshy (Meta)
Gaussian processes at the Helm(holtz): A better way to model ocean currents paper posterRenato Berlinghieri (MIT); Tamara Broderick (MIT); Ryan Giordano (MIT); Tamay Ozgokmen (University of Miami); Kaushik Srinivasan (UCLA); Brian L Trippe (MIT); Junfei Xia (University of Miami)
Towards Improved Learning in Gaussian Processes: The Best of Two Worlds paper posterRui Li (Aalto University); ST John (Aalto University); Arno Solin (Aalto University)
HyperBO+: Pre-training a universal hierarchical Gaussian process prior for Bayesian optimization paper posterZhou Fan (Harvard University); Xinran Han (Harvard University); Zi Wang (Google)
Sequential Gaussian Processes for Online Learning of Nonstationary Functions paperMichael Zhang (University of Hong Kong); Bianca  M Dumitrascu (Princeton  University); Sinead A Williamson (UT Austin); Barbara Engelhardt (Princeton University)
Gaussian Process Thompson sampling for Bayesian optimization of dynamic masking-based language model pre-training paper posterIñigo Urteaga (Columbia University); Moulay Zaidane Draidia (Columbia University); Tomer Lancewicki (Walmart Global Tech); Shahram Khadivi (eBay, Inc.)
Gaussian Process Regression for In-vehicle Disconnect Clutch Transfer Function Development paper posterHuanyi Shui (Ford Motor Co.); Yijing Zhang (Ford Motor Co.); Deepthi Antony (Ford Motor Co.); Devesh Upadhyay (Ford Motor Co.); James McCallum (Ford Motor Co.); Yuji Fujii (Ford Motor Co.); Edward Dai (Ford Motor Co.)
Variational Inference for Extreme Spatio-Temporal Matrix Completion paper posterCharul Paliwal (IIIT Delhi); Pravesh Biyani (IIIT Delhi)
Preprocessing Data of Varying Trial Duration with Linear Time Warping to Extend on the Applicability of SNP-GPFA paper posterArjan Dhesi (University of Edinburgh); Arno Onken (University of Edinburgh)
Are All Training Data Useful? A Empirical Revisit of Subset Selection in Bayesian Optimization paper posterPeili Mao (University of Electronic Science and Technology of China); Ke Li (University of Exeter)
Non-Gaussian Process Regression paper posterYaman Kindap (University of Cambridge); Simon Godsill (University of Cambridge)
Surrogate-Assisted Evolutionary Multi-Objective Optimization for Hardware Design Space Exploration paper posterRenzhi Chen (National Innovative Institute of Defense Technology); Ke Li (University of Exeter)
Bayesian Spatial Clustered Regression for Count Value Data paperPeng Zhao (Texas A&M University); Hou-Cheng Yang (FDA); Dipak Dey (University of Connecticut); Guanyu Hu (University of Missouri)
Efficient Variational Gaussian Processes Initialization via Kernel-based Least Squares Fitting paper posterXinran Zhu (Cornell University); David Bindel (Cornell University); Jacob Gardner (University of Pennsylvania)
Variational Bayesian Inference and Learning for Continuous Switching Linear Dynamical Systems paper posterJack Goffinet (Duke University); David Carlson (Duke University)
Adaptive Experimentation at Scale paper posterEthan Che (Columbia University); Hongseok Namkoong (Ass Prof Columbia)
Preference-Aware Constrained Multi-Objective Bayesian Optimization paper posterAlaleh Ahmadianshalchi (Washington State University); Syrine Belakaria (Washington State university); Janardhan Rao Doppa (Washington State University)
Constraining Gaussian Processes to Systems of Linear Ordinary Differential Equations paper posterAndreas Besginow (Technische Hochschule Ostwestfalen-Lippe); Markus Lange-Hegermann (Technische Hochschule Ostwestfalen-Lippe)
Imputation and forecasting for Multi-Output Gaussian Process in Smart Grid paper posterJIANGJIAO XU (University of Exeter); Ke Li (University of Exeter)
Shaping of Magnetic Field Coils in Fusion Reactors using Bayesian Optimisation paper posterTimothy Nunn (UK Atomic Energy Authority); Vignesh Gopakumar (United Kingdom Atomic Energy Authority); Sebastien Kahn (UK Atomic Energy Authority)
Joint Point Process Model for Counterfactual Treatment--Outcome Trajectories Under Policy Interventions paper posterÇağlar HIZLI (Aalto University); ST John (Aalto University); Anne Juuti (University of Helsinki); Tuure Saarinen (University of Helsinki); Kirsi Pietiläinen (University of Helsinki); Pekka Marttinen (Aalto University)
PI is back! Switching Acquisition Functions in Bayesian Optimization paper posterCarolin Benjamins (Leibniz University Hanover); Elena Raponi (Technical University of Munich); Anja Jankovic (Sorbonne University); Koen van der Blom (Sorbonne University); Maria Laura Santoni (University of Camerino); Marius Lindauer (Leibniz University Hannover); Carola Doerr (Sorbonne University)
Actually Sparse Variational Gaussian Processes paper posterHarry J Cunningham (UCL); So Takao (UCL); Mark van der Wilk (Imperial College London); Marc Deisenroth (University College London)
Predicting Spatiotemporal Counts of Opioid-related Fatal Overdoses via Zero-Inflated Gaussian Processes paper posterKyle Heuton (Tufts University); Shikhar Shrestha (Tufts University); Thomas Stopka (Tufts University); Jennifer Pustz (Tufts University); Liping Liu (Tufts University); Michael C Hughes (Tufts University)
Expert Selection in Distributed Gaussian Processes: A Multi-label Classification Approach paper posterHamed Jalali (University of Tuebingen); Gjergji Kasneci (  University of Tuebingen)
Statistical Downscaling of Sea Surface Temperature Projections with a Multivariate Gaussian Process Model paper posterAyesha Ekanayaka (University of Cincinnati); Emily Kang (University of Cincinnati); Peter Kalmus (Jet Propulsion Laboratory, California Institute of Technology); Amy Braverman (Jet Propulsion Laboratory, California Institute of Technology)
Multi-Mean Gaussian Processes: A novel probabilistic framework for multi-correlated longitudinal data paper posterArthur Leroy (The University of Manchester); Mauricio A Álvarez (University of Manchester)
Spatiotemporal Residual Regularization with Kronecker Product Structure for Traffic Forecasting paperSeongjin Choi (McGill University); Nicolas Saunier (Polytechnique Montreal); Martin Trepanier (École Polytechnique de Montréal); Lijun Sun (McGill University)
Uncertainty Disentanglement with Non-stationary Heteroscedastic Gaussian Processes for Active Learning paper posterZeel B Patel (IIT Gandhinagar); Nipun Batra (IIT Gandhinagar); Kevin Murphy (Google)
Active Learning with Convolutional Gaussian Neural Processes for Environmental Sensor Placement paper posterTom R Andersson (British Antarctic Survey); Wessel Bruinsma (Cambridge); Stratis Markou (University of Cambridge); James R Requeima (University of Cambridge); Alejandro Coca-Castro (The Alan Turing Institute); Anna Vaughan (Univeristy of Cambridge); Anna-Louise Ellis (Met Office); Matthew Lazzara (University of Wisconsin-Madison); Daniel C. Jones (British Antarctic Survey); J. Scott Hosking (British Antarctic Survey); Richard E. Turner (University of Cambridge)
Random Features Approximation for Fast Data-Driven Control paper posterKimia Kazemian (Cornell University); Sarah Dean (Cornell University)
Deep Mahalanobis Gaussian Process paper posterDaniel A de Souza (University College London); Diego Mesquita (Getulio Vargas Foundation); César Lincoln C Mattos (Federal University of Ceará); João Paulo Gomes (Federal University of Ceará)
An Empirical Analysis of the Advantages of Finite vs.~Infinite Width Bayesian Neural Networks paper posterJiayu Yao (Harvard University); Yaniv Yacoby (Harvard University); Beau Coker (Harvard University); Weiwei Pan (Harvard University); Finale Doshi-Velez (Harvard)
Non-exchangeability in Infinite Switching Linear Dynamical Systems paperVictor Geadah (Princeton University); Jonathan W Pillow (Princeton University)
Posterior Consistency for Gaussian Process Surrogate Models with Generalized Observations paper posterRujian Chen (MIT); John Fisher (MIT)
Recommendations for Baselines and Benchmarking Approximate Gaussian Processes paper posterSebastian W Ober (University of Cambridge); David R Burt (Massachusetts Institute of Technology); Artem Artemev (Imperial College London); Mark van der Wilk (Imperial College London)
Challenges in Gaussian Processes for Non Intrusive Load Monitoring paperAadesh K Desai (IIT Gandhinagar); Gautam Prashant Vashishtha (IIT Gandhinagar); Zeel B Patel (IIT Gandhinagar); Nipun Batra (IIT Gandhinagar)