Logo
  • ML Training
  • Projects
  • Software
  • Publications
  • Past news
  • About us

Simulation-based inference for scientific discovery

Type
Workshop
Area
Training
Author

Alvaro Tejero-CanteroAlvaro Tejero-Cantero

Last edited time
Oct 7, 2022 12:36 PM

Online, 20, 21 and 22 September 2021, 9am - 5pm CEST.

👻 missed this workshop?

There will be future editions - stay tuned on Twitter or ...

... join our Zulip stream on simulation-based inference! We are building a community open to both simulation scientists and machine learning methods researchers.

image

Goal

You use simulation in physics, economics, archaeology and want to find the simulator parameters that best fits the observations?

We teach a practical simulation-based inference workshop to help you understand recent machine learning techniques and apply them to your problem.

Apply to learn, have fun, and participate in a supportive community! (applications closed @August 31, 2021).

Program (1)

Day 1
1.1 Simulators and inference1.1 Simulators and inference
1.1 Simulators and inference

Álvaro Tejero-Cantero

🎚️Introduction to sbi: simulation & inference

1.2 Practical: from ABC to SBI1.2
1.2 Practical: from ABC to SBI

Álvaro Tejero-Cantero

Notebook (GitHub)

1.3 Invited lecture: SBI in neuroscience1.3
1.3 Invited lecture: SBI in neuroscience

Pedro J. Gonçalves

PDF slides

Day 2
2.1 (Conditional) density estimation2.1 (Conditional) density estimation
2.1 (Conditional) density estimation

Michael Deistler

Notebook (GitHub)

2.2 Sequential neural posterion estimation explained2.2 Sequential neural posterion estimation explained
2.2 Sequential neural posterion estimation explained

Michael Deistler

Notebook (GitHub)

2.3 SNLE and SNRE2.3
2.3 SNLE and SNRE

David Greenberg

Notebook (GitHub)

Day 3
3.1 Benchmarking sbi3.1
3.1 Benchmarking sbi

Jan-Matthis Lückmann

PDF slides

3.2 Introduction to the sbi toolkit3.2 Introduction to the sbi toolkit
3.2 Introduction to the sbi toolkit

Jan Bölts

Notebook (GitHub)

3.3 Troubleshooting sbi3.3 Troubleshooting sbi
3.3 Troubleshooting sbi

Jan Bölts

Notebook (GitHub)

3.4 Bayesian workflow3.4
3.4 Bayesian workflow

Jan Bölts

Notebook (GitHub)

Extras (day 3)
3.5 Bayesian model comparison3.5 Bayesian model comparison
3.5 Bayesian model comparison

Jan Bölts

PDF slides

3.6 Bring your own simulator and apply sbi to your problem!3.6 Bring your own simulator and apply sbi to your problem!
3.6 Bring your own simulator and apply sbi to your problem!

Participant led!

N/A

Instructors

We are a group of advanced users, package developers, and sbi researchers.

Álvaro Tejero-Cantero
Álvaro Tejero-Cantero, mlcolab @ Tübingen University
Michael Deistler
Michael Deistler, mackelab @ Tübingen University
Jan Bölts
Jan Bölts, mackelab @ Tübingen University
Pedro J. Gonçalves
Pedro J. Gonçalves, mackelab @ caesar
Jan-Matthis Lückmann
Jan-Matthis Lückmann, mackelab @ Tübingen University
David Greenberg
David Greenberg, Helmholtz Zentrum Hereon

Organizers

Álvaro Tejero-Cantero
Álvaro Tejero-Cantero, mlcolab @ Tübingen University
Peter Steinbach
Peter Steinbach, Helmholtz-Zentrum Dresden-Rossendorf
Alex Gessner
Alex Gessner, mlcolab @ Tübingen University
Stefan Wezel
Stefan Wezel, mlcolab @ Tübingen University
Daniela Huppenkothen
Daniela Huppenkothen, SRON Netherlands Institute for Space Research
Elena Sizana, mlcolab @ Tübingen University
Elena Sizana, mlcolab @ Tübingen University
image
image
image
Cluster of excellence Machine Learning, Uni Tübingen
Cluster of excellence Machine Learning, Uni Tübingen

Code of conduct

To ensure a comfortable, supportive, and respectful environment, the workshop was run under a code of conduct. You can download it below.

SBI_Workshop-Code_of_Conduct.pdf37.9KB

Any other Questions?

Our Zulip stream on simulation-based inference is open to both simulation scientists and machine learning methods researchers. Please write to Elena Sizana if you have trouble joining.

‣
Subpages
Logo

MLColab  ⊂

Cluster ML in Science  ⊂

University of Tübingen

::: Visit us

🄯 ml colab team, licensed cc-by-sa except where noted.

XGitHubMastodon