- Видео 105
- Просмотров 46 893
Machine Learning and Dynamical Systems Seminar
Великобритания
Добавлен 27 окт 2020
Seminars on Machine Learning and Dynamical Systems organized for the Special Interest Group (SIG) on Machine Learning and Dynamical Systems hosted by the Alan Turing Institute, cf. sites.google.com/site/boumedienehamzi/machine-learning-and-dynamical-systems-seminar for details about the SIG and how to join the mailinglist.
See also ruclips.net/video/YTIIYyUeNxk/видео.html for an introductory video to the SIG.
See also ruclips.net/video/YTIIYyUeNxk/видео.html for an introductory video to the SIG.
Mark Bishop, "Deep Stupidity: A Provocation on the Things LLMs Can and Cannot Do."
Date: Thursday, June 27, 2024
Speaker: Prof J. Mark Bishop
Title: Deep Stupidity: A Provocation on the Things LLMs Can and Cannot Do
Abstract: Artificial Neural Networks (ANNs) have soared to heights of "grandmaster" and "super-human" performance in games ranging from Go to Starcraft, heralding a new era of applications in business and becoming synonymous with the AI brand. However, when these systems falter-resulting in anything from autonomous vehicle accidents to biased automated processes-the repercussions quickly dominate headlines, raising questions about the reliability and understanding of AI.
Judea Pearl argued that the fundamental flaw is that these systems merely "fit curves" witho...
Speaker: Prof J. Mark Bishop
Title: Deep Stupidity: A Provocation on the Things LLMs Can and Cannot Do
Abstract: Artificial Neural Networks (ANNs) have soared to heights of "grandmaster" and "super-human" performance in games ranging from Go to Starcraft, heralding a new era of applications in business and becoming synonymous with the AI brand. However, when these systems falter-resulting in anything from autonomous vehicle accidents to biased automated processes-the repercussions quickly dominate headlines, raising questions about the reliability and understanding of AI.
Judea Pearl argued that the fundamental flaw is that these systems merely "fit curves" witho...
Просмотров: 998
Видео
Linfeng Wang: Measure transport with kernel mean embeddings
Просмотров 1252 месяца назад
Date: 20 June 2024 Speaker: Linfeng Wang Title: Measure transport with kernel mean embeddings Abstract: Kalman filters constitute a scalable and robust methodology for approximate Bayesian inference, matching first and second order moments of the target posterior. To improve the accuracy in nonlinear and non-Gaussian settings, we extend this principle to include more or different characteristic...
Liv Vage: Graph neural nets and reinforcement learning for particle tracking at the LHC
Просмотров 1952 месяца назад
Date: 13 June 2024 Speaker: Liv Vage Title: Graph neural nets and reinforcement learning for particle tracking at the LHC Abstract: The Large Hadron Collider (LHC) at CERN in Switzerland produces data at a rate on the order of tens of petabytes per second, comparable with the data size and processing requirements of Google Cloud. Not all this data can be stored, and the main experiments at CERN...
Florian Schaefer: Statistical Inference and PDEs: From operator learning to shock capturing
Просмотров 1722 месяца назад
Date: 06 June 2024 Speaker: Florian Schaefer Title: Statistical Inference and PDEs: From operator learning to shock capturing Abstract: The guiding theme of this talk is the interplay of statistics and partial differential equations (PDEs), in both new and classical problems. The first part of the talk is concerned with the learning of solution operators of PDEs, which has recently seen intense...
Boya Hou: Nonparametric Compressed Learning of Dynamical Systems
Просмотров 1352 месяца назад
Date: 30 May 2024 Speaker: Boya Hou Title: Nonparametric Compressed Learning of Dynamical Systems Abstract: The mature field of systems theory has enabled the success of model-based decision-making. Model identification typically requires fitting parametric models to data from interaction with the environments. In this talk, I will discuss an operator-theoretic approach to learning compressed r...
Matthew Colbrook: The Hitchhiker's Guide to the DMD Multiverse
Просмотров 1603 месяца назад
Date: 23 May 2024 Speaker: Matthew Colbrook Title: The Hitchhiker's Guide to the DMD Multiverse Slides: www.damtp.cam.ac.uk/user/mjc249/talks/rigged_DMD_turing.pdf Abstract: We introduce the Rigged Dynamic Mode Decomposition (Rigged DMD) algorithm, which computes generalized eigenfunction decompositions of Koopman operators. By considering the evolution of observables, Koopman operators transfo...
Lorenz Richter: An Optimal Control Perspective on Diffusion-Based Generative Modeling
Просмотров 1,3 тыс.3 месяца назад
Date: 16 May 2024 Speaker: Lorenz Richter Title: An Optimal Control Perspective on Diffusion-Based Generative Modeling Leading to Robust Numerical Methods Abstract: This seminar will delve into the intersection of generative modeling via Stochastic Differential Equations (SDEs) and three pivotal areas of mathematics: stochastic optimal control, Partial Differential Equations (PDEs), and path sp...
Nathan Doumèche: Physics-Informed Machine Learning as a Kernel Method
Просмотров 3043 месяца назад
Speaker: Nathan Doumèche Date: 09 May 2024 Title: Physics-Informed Machine Learning as a Kernel Method Abstract: Physics-informed machine learning merges the expressiveness of data-driven models with the interpretability of physical theories. This seminar explores a general regression problem where the empirical risk is augmented by a partial differential equation to enhance the model's adheren...
Nicolas Boulle: Elliptic PDE learning is provably data-efficient
Просмотров 2263 месяца назад
Speaker: Nicolas Boulle Date: 02 May 2024 Title: Elliptic PDE learning is provably data-efficient Abstract: PDE learning is an emerging field at the intersection of machine learning, physics, and mathematics, that aims to discover properties of unknown physical systems from experimental data. Popular techniques exploit the approximation power of deep learning to learn solution operators, ...
Yuanzhao Zhang: Catch-22s of reservoir computing
Просмотров 5114 месяца назад
Title: Catch-22s of reservoir computing Speaker: Yuanzhao Zhang Abstract: Reservoir Computing (RC) is a simple and efficient model-free framework for forecasting the behavior of nonlinear dynamical systems from data. Here, we show that there exist commonly-studied systems for which leading RC frameworks struggle to learn the dynamics unless key information about the underlying system is already...
Petar Veličković: Graph Deep Learning: Monoids and time, Embracing asynchrony in (G)NNs
Просмотров 3214 месяца назад
Speaker: Petar Veličković Date: 18 April 2024 Title: Graph Deep Learning: Monoids and time, Embracing asynchrony in (G)NNs Abstract: Virtually all present graph neural network (GNN) architectures blur the distinction between the definition and invocation of the message function, forcing a node to send messages to its neighbours at every layer, synchronously. When applying GNNs to learn to execu...
B. Hamzi: On Bridging ML, Dynamical Systems, & Algorithmic Info. Th. Via SKFs and PDE Simplification
Просмотров 2334 месяца назад
Title: Bridging Machine Learning, Dynamical Systems, and Algorithmic Information Theory: Insights from Sparse Kernel Flows and PDE Simplification Abstract: This presentation delves into the intersection of Machine Learning, Dynamical Systems, and Algorithmic Information Theory (AIT), exploring the connections between these areas. In the first part, we focus on Machine Learning and the problem o...
Tobias Schröder: Energy Discrepancy: Training of Energy-Based Models without Scores or MCMC
Просмотров 1936 месяцев назад
Speaker: Tobias Schröder Date: 08 Feb. 2024 Title: Energy Discrepancy: Training of Energy-Based Models without Scores or MCMC ABSTRACT: Energy-Based Models (EBMs) are probabilistic models inspired from statistical physics that allow estimating the data generating distribution via deep neural networks. As such, energy-based models are a powerful approach to generative modelling with applications...
Shi Jin: Dimension Lifting for Quantum Computation of Partial Differential Eqns and Related Problems
Просмотров 1046 месяцев назад
Date and Time: Thursday, February 1st, at 5:00 PM UK time Speaker: Shi Jin, Shanghai Jiao Tong University Title: Dimension Lifting for Quantum Computation of Partial Differential Equations and Related Problems Abstract: Quantum computers hold the potential to achieve remarkable computational speed-ups, from algebraic to exponential, when compared to their classical counterparts. They have the p...
Cristina Cipriani: A mean-field optimal control approach for the training of NeurODEs & AutoencODEs.
Просмотров 2777 месяцев назад
Speaker: Cristina Cipriani. Date: 18 January 2024 Title: A mean-field optimal control approach for the training of NeurODEs & AutoencODEs. Abstract: NeurODEs are a special type of neural networks that incorporate shortcut connections, enabling their training to be interpreted as an optimal control problem. Our work encompasses two main aspects. Firstly, we consider the mean-field formulation of...
Islem Rekik: The landscape of generative GNNs in network neuroscience
Просмотров 1827 месяцев назад
Islem Rekik: The landscape of generative GNNs in network neuroscience
Yoshito Hirata: Unified time series analysis for nonlinear deterministic/stochastic systems
Просмотров 3468 месяцев назад
Yoshito Hirata: Unified time series analysis for nonlinear deterministic/stochastic systems
Daniel Wilczak: Recent advances in rigorous computation of Poincaré maps
Просмотров 1108 месяцев назад
Daniel Wilczak: Recent advances in rigorous computation of Poincaré maps
Stefan Klus: Kernel based approximation of the Koopman generator and Schrödinger operator
Просмотров 2288 месяцев назад
Stefan Klus: Kernel based approximation of the Koopman generator and Schrödinger operator
P.-C. Aubin: The reproducing kernels underlying LQ control and Kalman filtering, and their duality
Просмотров 1418 месяцев назад
P.-C. Aubin: The reproducing kernels underlying LQ control and Kalman filtering, and their duality
Mengjia Xu: TransformerG2G: Adaptive time-stepping for learning temporal graph embeddings
Просмотров 1558 месяцев назад
Mengjia Xu: TransformerG2G: Adaptive time-stepping for learning temporal graph embeddings
F. Ferrini: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (4/4)
Просмотров 1 тыс.9 месяцев назад
F. Ferrini: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (4/4)
S. Azzolin: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (3/4)
Просмотров 3319 месяцев назад
S. Azzolin: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (3/4)
A. Longa: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (2/4)
Просмотров 9319 месяцев назад
A. Longa: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (2/4)
G. Santin: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (1/4)
Просмотров 1 тыс.9 месяцев назад
G. Santin: A hands-on introduction to Graph Deep Learning, with examples in PyTorch Geometric (1/4)
Bharath Sriperumbudur: Johnson & Lindenstrauss meet Hilbert at a Kernel
Просмотров 20710 месяцев назад
Bharath Sriperumbudur: Johnson & Lindenstrauss meet Hilbert at a Kernel
Nathan Kutz: The future of governing equations
Просмотров 1,5 тыс.10 месяцев назад
Nathan Kutz: The future of governing equations
B. Hamzi: Kernel Flows and Kernel Mode Decomposition for learning dynamical systems from data
Просмотров 204Год назад
B. Hamzi: Kernel Flows and Kernel Mode Decomposition for learning dynamical systems from data
Saurabh Malani: Identification with partial information
Просмотров 138Год назад
Saurabh Malani: Identification with partial information
Aamal Hussain: Session 7 of the reading group on Dynamics of Games
Просмотров 107Год назад
Aamal Hussain: Session 7 of the reading group on Dynamics of Games