Výsledky vyhledávání - "Tailor, Dharmesh"

Report

Learning to Defer to a Population: A Meta-Learning Approach

Autor: Tailor, Dharmesh, Patra, Aditya, Verma, Rajeev, Manggala, Putra, Nalisnick, Eric

The learning to defer (L2D) framework allows autonomous systems to be safe and robust by allocating difficult decisions to a human expert. All existing work on L2D assumes that each expert is well-identified, and if any expert were to change, the sys

Externí odkaz: http://arxiv.org/abs/2403.02683

Zobrazit plný text záznamu

Report

The Memory Perturbation Equation: Understanding Model's Sensitivity to Data

Autor: Nickl, Peter, Xu, Lu, Tailor, Dharmesh, Möllenhoff, Thomas, Khan, Mohammad Emtiyaz

Understanding model's sensitivity to its training data is crucial but can also be challenging and costly, especially during training. To simplify such issues, we present the Memory-Perturbation Equation (MPE) which relates model's sensitivity to pert

Externí odkaz: http://arxiv.org/abs/2310.19273

Zobrazit plný text záznamu

Report

Exploiting Inferential Structure in Neural Processes

Autor: Tailor, Dharmesh, Khan, Mohammad Emtiyaz, Nalisnick, Eric

Neural Processes (NPs) are appealing due to their ability to perform fast adaptation based on a context set. This set is encoded by a latent variable, which is often assumed to follow a simple distribution. However, in real-word settings, the context

Externí odkaz: http://arxiv.org/abs/2306.15169

Zobrazit plný text záznamu

Report

Learning the optimal state-feedback via supervised imitation learning

Autor: Tailor, Dharmesh, Izzo, Dario

Imitation learning is a control design paradigm that seeks to learn a control policy reproducing demonstrations from expert agents. By substituting expert demonstrations for optimal behaviours, the same paradigm leads to the design of control policie

Externí odkaz: http://arxiv.org/abs/1901.02369

Zobrazit plný text záznamu

Report

On the stability analysis of deep neural network representations of an optimal state-feedback

Autor: Izzo, Dario, Tailor, Dharmesh, Vasileiou, Thomas

Recent work have shown how the optimal state-feedback, obtained as the solution to the Hamilton-Jacobi-Bellman equations, can be approximated for several nonlinear, deterministic systems by deep neural networks. When imitation (supervised) learning i

Externí odkaz: http://arxiv.org/abs/1812.02532

Zobrazit plný text záznamu

Report

Machine learning and evolutionary techniques in interplanetary trajectory design

Autor: Izzo, Dario, Sprague, Christopher, Tailor, Dharmesh

After providing a brief historical overview on the synergies between artificial intelligence research, in the areas of evolutionary computations and machine learning, and the optimal design of interplanetary trajectories, we propose and study the use

Externí odkaz: http://arxiv.org/abs/1802.00180

Zobrazit plný text záznamu

Akademický článek

Unconscious Biases in Neural Populations Coding Multiple Stimuli.

Autor: Keemink, Sander W.¹ swkeemink@scimail.eu, Tailor, Dharmesh V.² dharmesh.tailor@live.co.uk, van Rossum, Mark C. W.³ mark.vanrossum@nottingham.ac.uk

Publikováno v: Neural Computation. Dec2018, Vol. 30 Issue 12, p3168-3188. 21p. 5 Graphs.