Aapo Hyvärinen

Menu

New:

Arxiv: ICE-BeeM: Identifiable Conditional Energy-Based Deep Models

Arxiv: Relative gradient optimization of the Jacobian term in unsupervised deep learning

Arxiv: Independent innovation analysis for nonlinear vector autoregressive process

Arxiv: Modeling Shared Responses in Neuroimaging Studies through MultiView ICA

NeuroImage: Nonlinear ICA of fMRI reveals primitive temporal structures linked to rest, task, and behavioral traits

UAI2020: Hidden Markov Nonlinear ICA: Unsupervised Learning from Nonstationary Time Series

UAI2020: Robust contrastive learning and nonlinear ICA in the presence of outliers

AISTATS2020: Variational Autoencoders and Nonlinear ICA: A Unifying Framework

Software

Deep unsupervised learning

[NEW!]  Code for Time-Contrastive Learning

Linear unsupervised learning

FastICA: Fast Independent Component Analysis

ICASSO: Analyzing reliability of independent components

LiNGAM: Causal discovery based on non-Gaussianity

Natural image statistics / visual modelling

Natural Image Statistics package (code for the book);
alternatively the imageica package

Neuroimaging data analysis

SpeDeBox: Decoding EEG/MEG using spectral infomation

OCF: Analysing variability (nonstationarity) of connectivity

ISCTEST: Testing independent components

Fourier-ICA: Improved ICA by time-frequency transforms

[Most were programmed by others, but implementing algorithms I have (co-)developed]