Neural Net Learning as Functional Kernel Gradient Descent (ft. Arthur Jacot)
In the functional space, and with the right kernel to compare functions, neural net learning can actually be regarded as a nice gradient descent, which is even convex for some common loss functions, as discussed by Arthur Jacot, PhD candidate in mathematics at EPFL.
https://people.epfl.ch/arthur.jacot
Check Arthur's 2018 NeurIPS paper on the neural tangent kernel
https://www.youtube.com/watch?v=raT2ECrvbag
https://arxiv.org/abs/1806.07572
Видео Neural Net Learning as Functional Kernel Gradient Descent (ft. Arthur Jacot) канала ZettaBytes, EPFL
https://people.epfl.ch/arthur.jacot
Check Arthur's 2018 NeurIPS paper on the neural tangent kernel
https://www.youtube.com/watch?v=raT2ECrvbag
https://arxiv.org/abs/1806.07572
Видео Neural Net Learning as Functional Kernel Gradient Descent (ft. Arthur Jacot) канала ZettaBytes, EPFL
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
![Teaching a machine learning algorithm better | Farnood Salehi](https://i.ytimg.com/vi/VB2lMCdKPlg/default.jpg)
![Concurrent Algorithms with Rachid Guerraoui](https://i.ytimg.com/vi/SSPx2UOqJwA/default.jpg)
![How Miners Secure Bitcoin & Blockchains (ft. Hamza, Pavlovic & Wang)](https://i.ytimg.com/vi/pftrbzWM33Q/default.jpg)
![Federated learning (ft. Sai Praneeth Reddy Karimireddy)](https://i.ytimg.com/vi/zxzBJSfJa4Q/default.jpg)
![Bitcoin (ft. Rachid Guerraoui & Jad Hamza)](https://i.ytimg.com/vi/QmgVx27nA0A/default.jpg)
![Enter the Hydra: Towards Principled Bug Bounties and Exploit-Resistant Smart Contracts | Ari Juels](https://i.ytimg.com/vi/WGuu6BckaVQ/default.jpg)
![Cloud Computing (ft. Babak Falsafi)](https://i.ytimg.com/vi/nfocd5TDXs8/default.jpg)
![Secure Microarchitectural Design | Atri Bhattacharyya](https://i.ytimg.com/vi/PZaaNvtA0D4/default.jpg)
![Misunderstanding Blockchain (ft. Adi Seredinschi, Matej Pavlovic and Rachid Guerraoui)](https://i.ytimg.com/vi/hLO888jgbq4/default.jpg)
![The Big Data Setup of the Human Brain Project (ft. Anastasia Ailamaki)](https://i.ytimg.com/vi/3rvObjC3DBE/default.jpg)
![Entropy as a Fundamental Compression Limit (ft. Rüdiger Urbanke)](https://i.ytimg.com/vi/ETrpXTg4Xs4/default.jpg)
![Securify: Practical Security Analysis of Smart Contracts | Petar Tsankov](https://i.ytimg.com/vi/YNbj_JElzuc/default.jpg)
![The Potential of Data Visualization (ft. Kirell Benzi)](https://i.ytimg.com/vi/6NtTH8m45o4/default.jpg)
![Should we Trust Algorithms? (ft. Rachid Guerraoui)](https://i.ytimg.com/vi/XC69954gK6k/default.jpg)
![The Quest of an Optimal Algorithm (ft. Boaz Barak)](https://i.ytimg.com/vi/m4lC7XIo2lA/default.jpg)
![Shannon's Optimal Communication (ft. Rüdiger Urbanke)](https://i.ytimg.com/vi/av637PYUFP8/default.jpg)
![How to radically outperform Bitcoin with Quorums (ft. Matej Pavlovic)](https://i.ytimg.com/vi/ms_QXEQyJvA/default.jpg)
![The Evolving Architecture of the Web and its Impact on Security, Privacy and Latency | Nick Sullivan](https://i.ytimg.com/vi/Jm4cQpQTs7g/default.jpg)
![Achieving both Reliability and Learning in AI (ft. Boi Faltings)](https://i.ytimg.com/vi/O_h-uxeOXKA/default.jpg)
![Theoretical Guarantees for Clustering (ft. Ashkan Norouzi)](https://i.ytimg.com/vi/yLiS9-YUP18/default.jpg)
![Appearance Modelling and the Uncanny Valley (ft. Wenzel Jakob)](https://i.ytimg.com/vi/PBf8OpAjxTU/default.jpg)