Aapo Hyvärinen

Menu

New:

Arxiv: A noise-corrected Langevin algorithm and sampling by half-denoising

Front Mol BioSci: Towards Interpretable CryoEM: Disentangling Latent Spaces of Molecular Conformations

ICML2024: Causal Representation Learning Made Identifiable by Grouping of Observational Variables

AISTATS2024: Identifiable Feature Learning for Spatial Data with Nonlinear ICA

NeurIPS2023: Provable benefits of annealing for estimating normalizing constants

NeuroImage: Unsupervised representation learning of spontaneous MEG data with nonlinear ICA

NeuroAI Group at University of Helsinki

NeuroAI (earlier we used the term neuroinformatics) is widely defined as the cross-fertilization of information-processing and mathematical sciences on the one hand, and neural and cognitive sciences on the other.

Our group works on different aspects of NeuroAI related to machine learning. Some examples include:

• We develop the theory of statistical machine learning, typically unsupervised.
• We model the visual system in the brain by analyzing the statistical structure of the natural input images.
• We apply machine learning models on neuroimaging data, in particular MEG.
• We develop a NeuroAI approach to affective phenomena, in particular human suffering.

The group is located at the department of Computer Science of the University of Helsinki and its leader is Aapo Hyvärinen.