# jemdoc: menu{MENU}{talks.html}, showsource
= Nadav Cohen --- Selected Talks
[https://focm2023.pages.math.cnrs.fr/ *FoCM 2023*] ([https://maps.app.goo.gl/DPsVs3BRi8qmHbzi8 Paris, France]) \n
Invited talk \n
/What Makes Data Suitable for Deep Learning?/ \n
\[[https://cohennadav.github.io/files/focm23_slides.pdf slides]\]
[https://icml.cc/Conferences/2022 *ICML 2022*] ([https://goo.gl/maps/eDNknEVeEsSvjT227 Baltimore, MD, USA]) \n
Invited talk \n
/Continuous vs. Discrete Optimization of Deep Neural Networks/ \n
\[[https://icml.cc/virtual/2022/workshop/13452\#wse-detail-18570 video]|[https://cohennadav.github.io/files/icml22_slides.pdf slides]\]
[https://www.ipam.ucla.edu/programs/workshops/workshop-i-tensor-methods-and-their-applications-in-the-physical-and-data-sciences/ *Workshop on Tensor Methods and their Applications in the Physical and Data Sciences 2021*] (virtual) \n
Invited talk \n
/Implicit Regularization in Deep Learning: Lessons Learned from Matrix and Tensor Factorization/ \n
\[[https://www.youtube.com/watch?v=doW3JvBOvCI&t=0 video]|[https://cohennadav.github.io/files/tensorw21_slides.pdf slides]\]
[https://2020.imvc.co.il/ *IMVC 2020*] (virtual) \n
Invited keynote talk \n
/Practical Implications of Theoretical Deep Learning/ \n
\[[https://www.youtube.com/watch?v=k3DVZltNyEU&t=0 video]|[https://cohennadav.github.io/files/imvc20_slides.pdf slides]\]
[https://nips.cc/Conferences/2019 *NeurIPS 2019*] ([https://goo.gl/maps/mqBaWhLu8uD71swm7 Vancouver, Canada]) \n
Contributed talk \n
/Implicit Regularization in Deep Matrix Factorization/ \n
\[[https://slideslive.com/38921800?time=16m1s video]|[https://cohennadav.github.io/files/neurips19_slides.pdf slides]\]
*AI Week 2019* ([https://goo.gl/maps/C1Zp993CtnEdkSP27 Tel Aviv-Yafo, Israel]) \n
Invited talk \n
/Analyzing Optimization and Generalization in Deep Learning via Trajectories of Gradient Descent/ \n
\[[https://www.youtube.com/watch?v=ZPhXCrqHlY4 video]|[https://cohennadav.github.io/files/aiweek19_slides.pdf slides]\]
*CECAM Workshop on Quantum Computing and Quantum Chemistry 2019* ([https://goo.gl/maps/C1Zp993CtnEdkSP27 Tel Aviv-Yafo, Israel]) \n
Invited talk \n
/Expressiveness in Deep Learning via Quantum Entanglement/ \n
\[[https://cohennadav.github.io/files/cecamw19_slides.pdf slides]\]
[https://simons.berkeley.edu/workshops/dl2019-1 *Simons Institute Workshop on Frontiers of Deep Learning 2019*] ([https://goo.gl/maps/S7F17tu1Z8TYP8838 Berkeley, CA, USA]) \n
Invited talk \n
/Analyzing Optimization and Generalization in Deep Learning via Trajectories of Gradient Descent/ \n
\[[https://www.youtube.com/watch?v=Lmj2bU9MdwM&list=PLgKuh-lKre11ekU7g-Z_qsvjDD8cT-hi9&index=4&t=0s video]|[https://cohennadav.github.io/files/simonsw19_slides.pdf slides]\]
[https://icerm.brown.edu/programs/sp-s19/w1/ *ICERM Workshop on Theory and Practice in Machine Learning and Computer Vision 2019*] ([https://goo.gl/maps/FxCMzeSxSXEfmFcs7 Providence, RI, USA]) \n
Invited talk \n
/Analyzing Optimization in Deep Learning via Trajectories/ \n
\[[https://icerm.brown.edu/video_archive/?play=1826 video]|[https://cohennadav.github.io/files/icermw19_slides.pdf slides]\]
[https://icml.cc/Conferences/2018 *ICML 2018*] ([https://goo.gl/maps/W1bxyK2TEPScAnAK9 Stockholm, Sweden]) \n
Contributed talk \n
/On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization/ \n
\[[https://vimeo.com/287857368 video]|[https://cohennadav.github.io/files/icml18_slides.pdf slides]\]
[https://iclr.cc/Conferences/2018 *ICLR 2018*] ([https://goo.gl/maps/mqBaWhLu8uD71swm7 Vancouver, Canada]) \n
Contributed talk \n
/Boosting Dilated Convolutional Networks with Mixed Tensor Decompositions/ \n
\[[http://www.youtube.com/watch?v=CXLiEe83EQs&t=58m15s video]|[https://cohennadav.github.io/files/iclr18_slides.pdf slides]\]
[https://sites.google.com/site/princetondeepmath *Symposium on the Mathematical Theory of Deep Neural Networks 2018*] ([https://goo.gl/maps/kgXqNRPvx7bckzif9 Princeton, NJ, USA]) \n
Invited talk \n
/On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization/ \n
\[[https://www.youtube.com/watch?v=SQJAHxKs0jY&list=PLWQvhvMdDChyI5BdVbrthz5sIRTtqV6Jw&index=7&t=0s video]|[https://cohennadav.github.io/files/mathdnn18_slides.pdf slides]\]
*Symposium on Physics and Machine Learning 2017* ([https://goo.gl/maps/CFtBkJzTGvUfqKxD7 New York City, NY, USA]) \n
Invited talk \n
/Understanding Deep Learning via Physics: The Use of Quantum Entanglement for Studying the Inductive Bias of Convolutional Networks/ \n
\[[https://cohennadav.github.io/files/phyml17_slides.pdf slides]\]
[https://www.wias-berlin.de/workshops/DL17/ *Mathematics of Deep Learning Workshop 2017*] ([https://goo.gl/maps/AEVNkuT8JFiEnipL9 Berlin, Germany]) \n
Invited talk \n
/Expressiveness of Convolutional Networks via Hierarchical Tensor Decompositions/ \n
\[[https://cohennadav.github.io/files/mathdl17_slides.pdf slides]\]
*AAAI Spring Symposium Series 2017* ([https://goo.gl/maps/njok2dFrGrnbgJGB6 Palo Alto, CA, USA]) \n
Invited talk \n
/Expressive Efficiency and Inductive Bias of Convolutional Networks: Analysis \& Design via Hierarchical Tensor Decompositions/ \n
\[[https://cohennadav.github.io/files/aaaisss17_slides.pdf slides]\]
[http://cvpr2017.thecvf.com/ *CVPR 2017*] ([https://goo.gl/maps/DJeHki3pXrh3CxRp8 Honolulu, HI, USA]) \n
Invited talk \n
/Expressive Efficiency and Inductive Bias of Convolutional Networks: Analysis \& Design via Hierarchical Tensor Decompositions/ \n
\[[https://cohennadav.github.io/files/cvpr17_slides.pdf slides]\]
*GAMM 2017* ([https://goo.gl/maps/bQKNwQdSoXCfuXyb8 Weimar, Germany]) \n
Invited talk \n
/On the Expressive Power of Deep Learning: A Tensor Analysis/ \n
\[[https://cohennadav.github.io/files/gamm17_slides.pdf slides]\]
[https://nips.cc/Conferences/2016 *NeurIPS 2016*] ([https://goo.gl/maps/G6MHfZUgkpGs7gkf8 Barcelona, Spain]) \n
Invited talk \n
/Inductive Bias of Deep Convolutional Networks through Pooling Geometry/ \n
\[[https://cohennadav.github.io/files/nips16_slides.pdf slides]\]
[https://icml.cc/2016/index.html *ICML 2016*] ([https://goo.gl/maps/CFtBkJzTGvUfqKxD7 New York City, NY, USA]) \n
Contributed talk \n
/Convolutional Rectifier Networks as Generalized Tensor Decompositions/ \n
\[[http://techtalks.tv/talks/convolutional-rectifier-networks-as-generalized-tensor-decompositions/62515/ video]|[https://cohennadav.github.io/files/icml16_slides.pdf slides]\]
[https://www.learningtheory.org/colt2016/ *COLT 2016*] ([https://goo.gl/maps/CFtBkJzTGvUfqKxD7 New York City, NY, USA]) \n
Contributed talk \n
/On the Expressive Power of Deep Learning: A Tensor Analysis/ \n
\[[https://www.youtube.com/watch?v=-spX4OMYoyg video]|[https://cohennadav.github.io/files/colt16_slides.pdf slides]\]