Donders | M1.02

Plantage Muidergracht 12, 1018 TV Amsterdam

Jul 18 @ 09:00 UTC - Jul 18 @ 12:30 UTC

Public session

This workshop will provide an introduction to deep learning methods and architectures for efficient Bayesian inference with complex models. It will include a self-contained theoretical part and a practical part, focusing on the topics of posterior estimation, model comparison, likelihood estimation, and model misspecification. In the theoretical part, participants will learn about neural density estimation with normalizing flows, simulation-based optimization, embedding networks, sequential and amortized inference, as well as the rationale of principled Bayesian workflows. In the practical part, participants will apply existing software packages for neural Bayesian inference (e.g., BayesFlow, SBI) to build their own amortized Bayesian workflows. Participants are highly encouraged to "bring" their own models and ideas to the workshop. In preparation, please install the required software packages: (1) Simple installation (pip): https://bayesflow.org/index.html#installation or (2) full installation instructions (and creating a conda environment): https://bayesflow.org/installation.html

Jul 18 @ 14:00 UTC - Jul 18 @ 17:30 UTC

Public session

In this workshop, we will engage with participants in an effort to curate formal models in psychology. For this purpose, we introduce the PsychoModels database that is currently under development as a platform for researchers to find, use, or contribute data generative models. Included models are annotated with information such as the psychological context, the modelling framework used, data used to parameterize the model (if applicable), and descriptive overview of the objects and functions inside the model. This workshop aims to test and improve a prototype in a crowd-sourced manner, in order to improve the database and create a community around it. Similar efforts have led to thriving platforms in modelling communities in other scientific fields (such as BioModels for systems biology or CoMSES for computational social science) — and we believe that a comprehensive and well-indexed platform for computational models will be a valuable resource for psychology at large, and mathematical psychology in particular. Firstly, a curated collection ensures that included models are clearly annotated and presented in a similar way, making it easier to skim a model and grasp its content than current repositories allow for. Secondly, this similar presentation facilitates a common language around modelling that makes it easier to communicate with other researchers. Thirdly, indexing model content allows to introduce search options across the model database and facilitate model review efforts, and the reuse or improvements of models in the database. Lastly, easing access to existing models and surrounding the database with educational materials will allow researchers with less experience in formalising their research to learn from best-practice. During the workshop, participants will first be made familiar with the current functionality of the database by creating entries of their own computational models. Building on this, we aim to discuss improvements to the platform and solutions to potential problems. Finally, we will develop ideas for the subsequent promotion and structural integration into the psychological community. We envision PsychoModels as a comprehensive resource that is exhaustive in its content and intuitive and beneficial a pleasure to use. This workshop aims to a) introduce modellers to the platform in its current state and benefits of such a resource and b) develop it with a community focus in mind, to design the usability according to researchers' needs. In the end, participants will have contributed to a shared resource that can be used to advance research and education in mathematical psychology and beyond. Their contributions will be acknowledged on the website of the database and we will invite them for future collaboration.

Jul 19 @ 09:00 UTC - Jul 19 @ 11:00 UTC

In-person session

Jul 19 @ 11:20 UTC - Jul 19 @ 12:40 UTC

In-person session

Jul 19 @ 15:20 UTC - Jul 19 @ 17:00 UTC

In-person session

Jul 20 @ 09:00 UTC - Jul 20 @ 10:40 UTC

In-person session

Jul 20 @ 11:00 UTC - Jul 20 @ 12:40 UTC

In-person session

Jul 20 @ 15:20 UTC - Jul 20 @ 16:20 UTC

In-person session

Jul 21 @ 09:00 UTC - Jul 21 @ 10:40 UTC

In-person session

Jul 21 @ 11:00 UTC - Jul 21 @ 12:40 UTC

In-person session

Jul 21 @ 15:20 UTC - Jul 21 @ 17:00 UTC

In-person session