Great presentation by G. Lemaitre at the funconnect meeting at Neurospin: How large cohorts (~2500 people) boosts prediction accuracy.

https://paris-saclay-cds.github.io/autism_challenge

See http://martinos.org/mne/stable/ and http://martinos.org/mne/stable/whats_new.html#version-0-17
for a complete change list

Congrats to all the developers team !

We are happy to announce the release of version 0.5 of nilearn, released
on Friday by Kshitij!

The most exciting addition in this release is interactive plotting
that naturally embeds in the Jupyter notebook, to plot volumes, surfaces,
or connectomes:
http://nilearn.github.io/plotting/index.html#interactive-plotting

A detailed changelog can be found at:
http://nilearn.github.io/whats_new.html

For his “Non-linear auto-regressive models for cross-frequency coupling in neural time series” paper.

Tom is supervised by Yves Grenier and Alexandre Gramfort.

Congratulations, Tom !

Arthur is defending his  Ph.D. thesis on the 28th of September, at 2pm in the Talairach amphitheatre, at NeuroSpin.

My thesis is entitled ‘Learning Representations from Functional fMRI Data’, and its abstract follows. The jury is composed of

Pr. Jalal Fadili, CNRS / Ensicaen, Caen, France – reviewer
Pr. Bin Yu, University of California, Berkeley, USA – reviewer
Pr. Moritz Große-Wentrup, Max-Planck Institute, Tübingen, Germany
Pr. Aapo Hyvärinen, University College London, UK
Pr. Eric Moulines, École Polytechnique, Palaiseau, France
Pr. Bertrand Thirion, Inria, Saclay, France – director
Pr. Julien Mairal, Inria, Grenoble, France – supervisor
Pr. Gaël Varoquaux, Inria, Saclay, France – supervisor

Congratulations to Arthur !

PhD defense of Carole Lazarus on Sept 27th at Neurospin at 2PM.

Jury members are:

Congratulations to Carole !

Congratulations to the development team !

Please See scikit-learn.org/stable/whats_new.html.

In https://hal.archives-ouvertes.fr/hal-01868735 we show how to create deep video representations that match the activity in the human visual system and discover novel relationships between these “deep encodings” and the human visual system.

Presented at CNN 2018  by Hugo Richard.

“Toward a rigorous causal framework for brain mapping”, following our collaboration with @zdeborova

@ogrisel has been giving a great course on deep learning with Keras on June 28-29 at Polytechnique, Palaiseau, France. See https://m2dsupsdlclass.github.io/lectures-labs/