Lecture

Neural Tangent Kernel: Generalization in Deep Learning

Graph Chat

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

In course

PHYS-754: Lecture series on scientific machine learning

This lecture presents ongoing work on how scientific questions can be tackled using machine learning. Machine learning enables extracting knowledge from data computationally and in an automatized way. We will learn on examples how this is influencing the v ...

Description

This lecture explores the concept of the neural tangent kernel (NTK) in deep learning, showing how it converges to a constant value as the neural network width approaches infinity. It discusses the impact of network size on generalization, highlighting the role of random noise and fluctuations during training and initialization. The lecture also delves into the asymptotic behavior of generalization as the number of parameters increases, demonstrating the variance of the output and its relationship with network size. Various measures of generalization are analyzed, revealing patterns such as peaks and decays. The presentation concludes with insights on the behavior of wide neural networks under gradient descent and their evolution into linear models.

Instructors (10)

Michele Ceriotti

Michele Ceriotti received his Ph.D. in Physics from ETH Zürich in 2010. He spent three years in Oxford as a Junior Research Fellow at Merton College. Since 2013 he leads the laboratory for Computational Science and Modeling in the Institute of Materials at EPFL. His research revolves around the atomic-scale modelling of materials, based on the sampling of quantum and thermal fluctuations and on the use of machine learning to predict and rationalize structure-property relations. He has been awarded the IBM Research Forschungspreis in 2010, the Volker Heine Young Investigator Award in 2013, an ERC Starting Grant in 2016, and the IUPAP C10 Young Scientist Prize in 2018.

David Richard Harvey

Expertise CosmologyDark MatterGalaxy ClustersStrong Gravitational LensingWeak Gravitational Lensing I am an observational and theoretical astrophysicist at EPFL. My main area of research is dark matter and trying to understand its properties and dynamics. I love to use deep learning in my work and understand how fondation models and generative AI can help us understand the Universe. I am a member of Euclid and ARRAKIHS and the lead of the self interacting dark matter Key Project in Euclid. Outside of my r

Matthieu Wyart

Teaching & PhD Past EPFL PhD Students Riccardo Ravasio, Mario Geiger, Wencheng Ji, Leonardo Petrini, Umberto Maria Tomasini, Alessandro Favero Courses Systèmes complexes PHYS-343 This courses show how known (entropy, phase transitions) and new (glasses, percolation, avalanche-type response) physical concepts can be used in various fields (including Computer science, neuroscience, social sciences).

Anne-Florence Raphaëlle Bitbol

I was trained in physics, and always had a strong interest in biology. I am broadly interested in understanding biological phenomena in a quantitative way, through physical concepts and mathematical and computational tools. I investigate the relative impacts of optimization and historical contingency in biological evolution, both at the scale of protein sequences and at the scale of microbial populations.Here is my research group website.Curriculum vitae Here is my complete CV. Professional experience:Since

Official source

Related concepts (32)

Linear model

In statistics, the term linear model is used in different ways according to the context. The most common occurrence is in connection with regression models and the term is often taken as synonymous with linear regression model. However, the term is also used in time series analysis with a different meaning. In each case, the designation "linear" is used to identify a subclass of models for which substantial reduction in the complexity of the related statistical theory is possible.

Convolutional neural network

Convolutional neural network (CNN) is a regularized type of feed-forward neural network that learns feature engineering by itself via filters (or kernel) optimization. Vanishing gradients and exploding gradients, seen during backpropagation in earlier neural networks, are prevented by using regularized weights over fewer connections. For example, for each neuron in the fully-connected layer 10,000 weights would be required for processing an image sized 100 × 100 pixels.

Generalized linear model

In statistics, a generalized linear model (GLM) is a flexible generalization of ordinary linear regression. The GLM generalizes linear regression by allowing the linear model to be related to the response variable via a link function and by allowing the magnitude of the variance of each measurement to be a function of its predicted value. Generalized linear models were formulated by John Nelder and Robert Wedderburn as a way of unifying various other statistical models, including linear regression, logistic regression and Poisson regression.

Deep learning

Deep learning is part of a broader family of machine learning methods, which is based on artificial neural networks with representation learning. The adjective "deep" in deep learning refers to the use of multiple layers in the network. Methods used can be either supervised, semi-supervised or unsupervised.

Linear regression

In statistics, linear regression is a linear approach for modelling the relationship between a scalar response and one or more explanatory variables (also known as dependent and independent variables). The case of one explanatory variable is called simple linear regression; for more than one, the process is called multiple linear regression. This term is distinct from multivariate linear regression, where multiple correlated dependent variables are predicted, rather than a single scalar variable.

Related lectures (64)

Deep Learning: Data Representations and Neural Networks

Explores data representations, histograms, neural networks, and deep learning concepts.

Provable Benefits of Overparameterization in Model Compression

Explores the provable benefits of overparameterization in model compression, emphasizing the efficiency of deep neural networks and the importance of retraining for improved performance.

Deep Learning: Data Representations and Neural Networks

Covers data representations, Bag of Words, histograms, data pre-processing, and neural networks.

Machine Learning Review

Covers a review of machine learning concepts, including supervised learning, classification vs regression, linear models, kernel functions, support vector machines, dimensionality reduction, deep generative models, and cross-validation.

Introduction to Machine Learning: Supervised Learning