BISTOM

Bayesian Parameter Inference for Stochastic Models

Started
April 1, 2018
Status
Completed
Share this project

Abstract

For a multitude of different field sciences, determining the underlying mechanistic models is important in order to further our understanding. An accurate estimation of the parameters of such mechanistic models through data can be computationally prohibitive. This project exploits neural networks in order to learn minimal and near sufficient summary statistics as latent embeddings on simulated data for multiple stochastic models. This is corroborated with sharp model parameter posteriors observed through approximate Bayesian computation experiments.

People

Collaborators

SDSC Team:
Firat Ozdemir
Fernando Perez-Cruz

PI | Partners:

Eawag, Mathematical Methods in Environmental Research:

  • Dr. Carlo Albert

More info

ZHAW, Biomedical Simulation:

  • Dr. Simone Ulzega

More info

USI, Data Science Lab:

  • Prof. Dr. Antonietta Mira

More info

Paul Scherrer Institute:

  • Prof. Dr. Christian Rüegg

More info

description

Goal

Reliably estimating parameters of mechanistic models from data (Bayesian inference) is computationally expensive if (i) the data is big or (ii) the model is stochastic. Stochastic models are needed for reliable predictions.

Impact

The developed methodology can be applied in many fields of science and engineering, wherever system understanding (mechanistic model) needs to be combined with data, for advancing domain knowledge and making more reliable predictions.

Solution

Advances in algorithms as well as parallel computing infrastructure allow for Bayesian inference to be applied to a large class of stochastic models and to be scaled up to big data. We developed neural network based framework which can learn minimal and near sufficient summary statistics as latent embeddings on simulated data for multiple stochastic models. Experiments with approximate Bayesian computation yield sharp model parameter posteriors. For both stochastic models, developed solution finds near sufficient summary statistics.

Gallery

Annexe

Additional resources

Bibliography

Publications

Raynal, L.; Chen, S.; Mira, A.; Onnela, J. "Scalable Approximate Bayesian Computation for Growing Network Models via Extrapolated and Sampled Summaries" Bayesian Analysis 17 1 2022 View publication
Albert, C.; Ulzega, S.; Ozdemir, F.; Perez-Cruz, F.; Mira, A. "Learning Summary Statistics for Bayesian Inference with Autoencoders" SciPost Physics Core 5 3 43 2022 View publication
Guratinder, K.; Schmidt, M.; Walker, H. C.; Bewley, R.; Wörle, M.; Cabra, D.; Osorio, S. A.; Villalba, M.; Madsen, A. K.; Keller, L.; et al. "Magnetic correlations in the triangular antiferromagnet FeGa2S4" Physical Review B 104 6 64412 2021 View publication
Albert, C.; Ferriz-Mas, A.; Gaia, F.; Ulzega, S. "Can Stochastic Resonance Explain Recurrence of Grand Minima?" The Astrophysical Journal Letters 916 2 L9 2021 View publication
Allenspach, S.; Puphal, P.; Link, J.; Heinmaa, I.; Pomjakushina, E.; Krellner, C.; Lass, J.; Tucker, G. S.; Niedermayer, C.; Imajo, S.; et al. "Revealing three-dimensional quantum criticality by Sr substitution in Han purple" Physical Review Research 3 2 23177 2021 View publication
Warne, D. J.; Ebert, A.; Drovandi, C.; Hu, W.; Mira, A.; Mengersen, K. "Hindsight is 2020 vision: a characterisation of the global response to the COVID-19 pandemic" BMC Public Health 20 1 1868 2020 View publication

Related Pages

More projects

Merck: cancer care adverse events

Completed
Exploring how machine learning can help address cancer treatment side effects
Private sector

Èvokâ: Enhancing Fribourg's Archives

Completed
Making 10 TB of cultural heritage searchable and emotionally resonant
Public sector
Academia

ACROSS

In Progress
Actionable, Collaborative, Reusable, interOperable Sepsis System
No items found.

Dedgeflow

Completed
Using Data Science and Edge Computing for Debris Flow Warning
Academia

More projects

No results found.
There are no results with this criteria. Try changing your search.
No result

News

Latest news

La science des données au service de la neutralité carbone du canton : ENERBAT ouvre la voie
September 17, 2026

La science des données au service de la neutralité carbone du canton : ENERBAT ouvre la voie

La science des données au service de la neutralité carbone du canton : ENERBAT ouvre la voie

News release : ENERBAT, un outil fondé sur les données soutient la politique de rénovation des bâtiments du Canton de Vaud.
The SDSC to support 31 projects through its second National Calls
August 13, 2026

The SDSC to support 31 projects through its second National Calls

The SDSC to support 31 projects through its second National Calls

News Release: The SDSC has accepted 31 proposals for collaborative projects in its 2025 National Calls.
SDSC and PSI teams jointly explore the AI potential of large-scale infrastructures
July 8, 2026

SDSC and PSI teams jointly explore the AI potential of large-scale infrastructures

SDSC and PSI teams jointly explore the AI potential of large-scale infrastructures

Tailored community event at PSI in Villigen, Aargau, reveals machine learning and AI opportunities in accelerating scientific discovery

Contact us

Would you like to explore how we can work together?

Our team of data scientists and professionals looks forward to engaging in impactful, results-driven collaborations with you.