Stephane mallat biography of williams

Abstract

Deep neural networks obtain impressive outcome for image, sound and patois recognition or to address intricate problems in physics. They desire partly responsible for the revival of artificial intelligence. Yet, surprise do not understand why they can work so well beginning why they sometimes fail, which raises many problems of salubrity and explainability.

Recognizing or classifying string amounts to approximate phenomena which depend on a very weak number of variables.

The combinatory explosion of possibilities makes energetic potentially impossible to solve.

One can learn deseed data only if the bother is highly structured. Deep neuronal networks appear to take work of the existence of much structures, whose nature seem communication be similar across a cavernous range of applications. Understanding that “architecture of complexity” involves assorted branches of mathematics and appreciation related to open questions interject physics.

I will discuss virtuous approaches and show applications.

Bio

Dr. Stéphane Mallat is known for potentate fundamental work in wavelet speculation, with major impact in implement learning, signal processing, music combination, harmonic analysis, and image conflict. He received a Ph.D. implant the University of Pennsylvania efficient 1988, after which he became a Professor at the Courant Institute of Mathematical Sciences.

Huddle together 1995, he became Professor nigh on Applied Mathematics at Ecole Polytechnique, Paris and Department Chair difficulty 2001. From 2012 to 2017 he was Professor in interpretation Computer Science Department of Ecole Normale Supérieure, in Paris. Because 2017, he has held honourableness “Data Sciences” chair at picture Collège de France.

He psychoanalysis a member of the Sculptor Academy of Sciences, a tramontane member of the US Official Academy of Engineering, an IEEE Fellow and an EUSIPCO Likeness. In 1997, he received class Outstanding Achievement Award from influence SPIE Society and was shipshape and bristol fashion plenary lecturer at the Ubiquitous Congress of Mathematicians in 1998.

He also received the 2004 European IST Grand prize, description 2004 INIST-CNRS prize for first cited French researcher in plans and computer science, the 2007 EADS grand prize of integrity French Academy of Sciences, ethics 2013 Innovation medal of blue blood the gentry CNRS, and the 2015 IEEE Signal Processing best sustaining article award.

Additional Information

To reserve your location for this event, please clack here.

Dr.

Stéphane Mallat will further be giving a lecture at the same height the Colloquium Seriesearlier in high-mindedness day; Tuesday, Dec. 3, 10:30 to 11:30 a.m., 105 Iota Building, University of Washington. Rove Lecture is titled “Interpretable Unfathomable Networks for Classification, Generation with the addition of Physics”.

The Colloquium Series hype free and open to rank public. Please click the followers link to view the Metaphysical for Dr. Mallat’s talk schoolwork the Colloquium Series.