Introduction to deep learning

Deep learning is a new AI technique inspired by the biology of neural networks which has revolutionized fields such as voice recognition and computer vision and is starting to be used in a variety of areas. The availability of vast amounts of data and powerful GPU accelerators has revealed the potential of deep learning to match and exceed human capabilities in completing tasks such as the board game Go (AlphaGo), recognition of traffic signs, facial recognition, etc. Deep learning could prove highly useful in improving our understanding of biological systems. It is becoming one of the most powerful tools for studying life sciences.

Now that new algorithms have been developed to make deep learning even more powerful, it is important to introduce more scientists to the technique. Thanks to the extensive efforts of large-scale IT companies and the AI research community, building powerful deep learning or neural network models has never been simpler.

To promote the concepts of deep learning and other new AI techniques on the Institut Pasteur campus, a "Deep Learning Club" has been set up by scientists and PhD students from the Pasteur-Paris University (PPU) International Doctoral Program.

The Deep Learning Club has invited three scientists based on campus to explain the technique of deep learning and related applications during a seminar that will take place on Tuesday October 11, 2016, from 1 to 2.35pm, in the Monod lecture hall.

You can sign up for the club and the email list by contacting deep.learning@pasteur.fr.

Upcoming events organized by the club will include journal club sessions and workshops on applications of new deep neural networks and algorithms, software tools, etc. for biology research.