NDL April meeting and MILA

Hello everyone!! We are here with the consolidated details of the April meeting and some other exciting announcements!!

April Meeting

As decided, we will be having the meeting from 4 pm on April 25, where Alexia Jolicoeur-Martineau will present her work on Generative Adversarial Networks (GANs).😎

Please find below the abstract of her talk :-

"Generative adversarial networks (GANs) are a very popular recent approach to generate data from a unknown probability distribution. GANs have a very broad range of applications: generating images, sound, video, tabular data, imaging data, etc. 

In standard generative adversarial network (SGAN), the discriminator estimates the probability that the input data is real. The generator is trained to increase the probability that fake data is real. We argue that it should also simultaneously decrease the probability that real data is real. We show that this property can be induced by using a relativistic discriminator which estimate the probability that the given real data is more realistic than a randomly sampled fake data. We also present a variant in which the discriminator estimate the probability that the given real data is more realistic than fake data, on average. We generalize both approaches to non-standard GAN loss functions and we refer to them respectively as Relativistic GANs (RGANs) and Relativistic average GANs (RaGANs). Empirically, we observe that these approaches are significantly more stable and generate higher quality data samples than their non-relativistic counterparts."

GANs are an extremely interesting topic in the Deep Learning literature and we would really encourage you to attend the talk. (Spoiler alert - GANs can generate data from noise!!). I am sure we will be having stimulating discussions on GANs and how we could leverage them in neuroscience for data augmentation. Speaking of discussions, we will not be having a video conference with other sites this time. 😞

We would still be having a live streaming link, that will be shared to you shortly before the meeting. In case you want to join the discussions and will not be able to make it to the MNI, shoot me an email and we can organise Skype calls (open to suggestions for this) for questions and comments with specific sites. 

Venue: de Grandpré Communications Centre, MNI, Neuro, McGill

Data: 25th April 2019

Time: 4:00 - 5:00 pm

Link to RSVP: http://ndl3.rsvpify.com

**May Meeting** Time is running out to suggest papers for the May meeting. We have received a few suggestions but if you want to discuss (or think the group can help in detailed understanding of) a paper, please fill out this form: https://forms.gle/ryAByaMpmWQgVie77

We are looking forward to your responses!! 😃

Neuroscience and AI reading/study group at MILA

One of our members, Raymond Chua (raymond.chua@mail.mcgill.ca), has created a new reading/study group at Mila, for the AI researchers to gain exposure to Neuroscience. The main focus will be on topics on the intersection between AI and Neuroscience.

Those interested are encouraged to send him an email so that he could bring you to the room.  The first meeting will be held on 19 April (Friday) 9am to 10am at Mila. Room: A14.

Agenda: Introduction and Paper Discussion 

Paper: Neuroscience-Inspired Artificial Intelligence

https://www.cell.com/neuron/fulltext/S0896-6273(17)30509-3 Address of Mila: 6666 St Urbain St, Montreal, QC H2S 3H1

==================================================

As always let us know if you have any questions, concerns, feedback, etc.