StanCon 2020

StanCon 2020

StanCon 2020. A 24h Global Event.

StanCon Thursday, 13 August 2020.

The conference was a 24-hour event with three main sessions spanning across different time zones (British Summer Time, Eastern Time and Pacific Time).


The sessions were recorded and uploaded to YouTube. The plenary talks and the discussions can be found in the recording for the live sessions.

The abstracts to the talks are included in the video descriptions.

Recorded StanCon 2020 Videos on YouTube


Session 1

08:30 03:30 00:30 Virtual coffee
09:00 04:00 01:00 Live welcome from StanCon UK Committee
09:04 04:04 01:04 6 x 1 minute pitches
09:10 04:10 01:10 Plenary: Hierarchical Models for Covid -
      identifying effects of lockdown and an R package,
      Seth Flaxman, Imperial College, London
09:50 04:50 01:50 Discussion
10:00 05:00 02:00 6 x 1 minute pitches, followed by 9 minute discussions:
10:06 05:06 02:06 Hierarchical Bayes and logistic growth: Predicting diagnosed
      COVID19 cases and the corresponding burden on health care
      systems, Brynjólfur Gauti Jónsson, Birgir Hrafnkelsson,
      Jón Magnús Jónsson, Jóhanna Jakobsdóttir,
      Bergdís Björg Sigurjónsdóttir,Þórarinn Jónmundsson,
      Sigríður Haraldsdóttir, Unnur A. Valdimarsdóttir,
      Thor Aspelund, Directorate of Health (Iceland),
      Center of Public Health Sciences (University of Iceland),
      Department of Mathematics (University of Iceland),
      Icelandic Heart Association, National University Hospital of Iceland
10:15 05:15 02:15 Process fault detection using Stan, Jerzy Baranowski,
      Waldemar Bauer, Rafał Mularczyk, Bartłomiej Gondek,
      AGH University of Science & Technology, Kraków, Poland
10:24 05:24 02:24 Failure prediction in hierarchical equipment system:
      spline fitting naval ship failure, Hyunji Moon, Jinwoo Choi,
      Hyeonseop Lee, Seoul National University
10:33 05:33 02:33 baggr: a new meta-analysis package using Stan,
      Witold Wiecek and Rachael Meager, London School of Economics
10:42 05:42 02:42 Towards an Interface for Streaming Stan,
      Simon Maskell and Alessandro Varsi, University of Liverpool
10:51 05:51 02:51 Enforcing stationarity through the prior in vector autoregressions,
      Sarah Heaps, Newcastle University
11:00 06:00 03:00 Virtual coffee and poster session
13:30 08:30 05:30 End of session

Session 2

16:30 11:30 08:30 Virtual coffee
17:00 12:00 09:00 Live welcome from StanCon UK + US Committee
17:10 12:10 09:10 Code of Conduct working group, Steve Bronder and Lauren Kennedy
17:14 12:14 09:14 Stan Governing Body update, Jonah Gabry
17:20 12:20 09:20 6 x 1 minute pitches
17:32 12:32 09:32 Developer talk 1: posteriordb: a database with data, models and
      posteriors, Aki Vehtari, Måns Magnusson, Aalto University
17:39 12:39 09:39 Developer talk 2: Scaling Stan’s performance with reduce_sum in practice,
      Sebastian Weber, Ben Bales, Steve Bronder, Mitzi Morris,
      Rok Češnovar, Imperial College of London COVID-19 team,
      Novartis Pharma AG, Basel, Switzerland, Columbia University,
      University of Ljubljana
17:46 12:46 09:46 Developer talk 3: The State of GPU Computation Support for Stan,
      Steve Bronder, Erik Štrumbelj, Rok Češnovar, Davor Sluga,
      Jure Demšar, Tadej Ciglarič, Sean Talts, Columbia University,
      University of Ljubljana
17:53 12:53 09:53 Developer talk 4: Approximate Bayesian inference for latent Gaussian
      models in Stan, Charles Margossian, Aki Vehtari, Daniel Simpson,
      Raj Agrawal, Columbia University, Department of Statistics;
      Aalto University, Department of Computer Science;
      University of Toronto, Department of Statistical Sciences;
      Massachusetts Institute of Technology, CSAIL
      video, code and notebook
18:00 13:00 10:00 Plenary: Multi-Source Information Modeling, Curation, and
      Fusion Enabling Transdisciplinary Decision-Making: A Case for Space!
      Moriba Jah, Oden Institute for Computational Engineering & Sciences
18:50 13:50 10:50 Discussion
19:00 14:00 11:00 6 x 1 minute pitches, followed by 9 minute discussions
19:06 14:06 11:06 Using Stan for Spatial Smoothing of Regression Parameters,
      Wade Brorsen, Bart Niyizibi, Davood Pourvina, Eunchun Park,
      Oklahoma State University
19:15 14:15 11:15 Multivariate Synthetic Control to Estimate Missing Data in
      Hierarchically Correlated Series, Sean Pinkney, Tyler Morrison,
      Gaurav Shahane, Comscore, Inc.
19:24 14:24 11:24 Bayesian Causal Effect Estimation with Stan: Parametric and
      Nonparametric Approaches, Arman Oganisian, Jason A. Roy,
      University of Pennsylvania
19:33 14:33 11:33 Building effective uncertainty visualizations with Stan/brms,
      tidybayes, and ggdist, Matthew Kay, University of Michigan
19:42 14:42 11:42 API for iterative interrogation of stanfit objects in R,
      Jonathan Sidi, Associate Director of Modeling and Simulation at
      Sage Therapeutics
19:51 14:51 11:51 Bayesian Inference without Probability Density Functions,
      Ben Goodrich, Columbia University
20:00 15:00 12:00 Virtual coffee and poster session
21:30 16:30 13:30 End of session

Session 3

00:30 19:30 16:30 Virtual coffee
01:00 20:00 17:00 Live welcome from StanCon Committee
01:04 20:04 17:04 6 x 1 minute pitches
01:10 20:10 17:10 Plenary: STAN and US Politics, David Shor
01:50 20:50 17:50 Discussion
02:00 21:00 18:00 6 x 1 minute pitches, followed by 9 minute discussions:
02:06 21:06 18:06 Spatial models for plant neighborhood dynamics in Stan,
      Cristina Barber, Andrii Zaiats, Cara Applestein, Trevor Caughlin,
      Boise State University
02:15 21:15 18:15 ​Stress-testing the Dawid–Skene model: Analysis of repeated
      categorical ratings, Damjan Vukcevic, Jeffrey Pullin, Lyle Gurrin,
      University of Melbourne
02:24 21:24 18:24 ArviZ, InferenceData, and NetCDF: A unified file format for Bayesians,
      Ravin Kumar, Oriol Abril-Pla, Osvaldo Martin, Piyush Gautam,
      Ari Hartikainen , Alexandre Andorra, ArviZ (all)
      video (English)
      video (Catalan)
      video (French)
      video (Finnish)
​02:33 21:33 18:33 Orbit: Probabilistic Forecast with Exponential Smoothing,
      Edwin Ng, Zhishi Wang, Steve Yang, Uber
02:42 21:42 18:42 Effortless frequentist covariances of posterior expectations in Stan,
      Ryan Giordano, Tamara Brodercik, MIT
02:51 21:51 18:51 Virtual coffee and poster session
05:30 00:30 21:30 End of session


We can’t do this without the support of generous sponsors!


  • Susana Marquez. The Rockefeller Foundation.
  • Daniel Lee. Generable.
  • Kelli Cassidy. The University of Liverpool.
  • Simon Maskell. The University of Liverpool.