[PDF][PDF] On semi-global optimal control problems and their applications in machine learning
M Oster - 2023 - depositonce.tu-berlin.de
Semi-global optimal control problems enable us to use open-loop-like methods to
understand optimal feedback synthesis for an averaged optimal control problem. In this …
understand optimal feedback synthesis for an averaged optimal control problem. In this …
Global optimality of softmax policy gradient with single hidden layer neural networks in the mean-field regime
We study the problem of policy optimization for infinite-horizon discounted Markov Decision
Processes with softmax policy and nonlinear function approximation trained with policy …
Processes with softmax policy and nonlinear function approximation trained with policy …
Mean-field langevin system, optimal control and deep neural networks
In this paper, we study a regularised relaxed optimal control problem and, in particular, we
are concerned with the case where the control variable is of large dimension. We introduce …
are concerned with the case where the control variable is of large dimension. We introduce …
[HTML][HTML] Deep learning and mean-field games: A stochastic optimal control perspective
LD Persio, M Garbelli - Symmetry, 2020 - mdpi.com
We provide a rigorous mathematical formulation of Deep Learning (DL) methodologies
through an in-depth analysis of the learning procedures characterizing Neural Network (NN) …
through an in-depth analysis of the learning procedures characterizing Neural Network (NN) …
Smooth approximation of feedback laws for infinite horizon control problems with non-smooth value functions
K Kunisch, D Vásquez-Varas - arXiv preprint arXiv:2312.11981, 2023 - arxiv.org
In this work the synthesis of approximate optimal and smooth feedback laws for infinite
horizon optimal control problems is addressed. In this regards, $ L^{p} $ type error bounds of …
horizon optimal control problems is addressed. In this regards, $ L^{p} $ type error bounds of …
A mean-field optimal control formulation of deep learning
Recent work linking deep neural networks and dynamical systems opened up new avenues
to analyze deep learning. In particular, it is observed that new insights can be obtained by …
to analyze deep learning. In particular, it is observed that new insights can be obtained by …
[PDF][PDF] A mean-field optimal control formulation of deep learning
Recent work linking deep neural networks and dynamical systems opened up new avenues
to analyze deep learning. In particular, it is observed that new insights can be obtained by …
to analyze deep learning. In particular, it is observed that new insights can be obtained by …
Gradient flows for regularized stochastic control problems
This paper studies stochastic control problems with the action space taken to be the space of
measures, regularized by the relative entropy. We identify suitable metric space on which we …
measures, regularized by the relative entropy. We identify suitable metric space on which we …
Actor-critic learning for mean-field control in continuous time
We study policy gradient for mean-field control in continuous time in a reinforcement
learning setting. By considering randomised policies with entropy regularisation, we derive a …
learning setting. By considering randomised policies with entropy regularisation, we derive a …
Global optimality in model predictive control via hidden invariant convexity
JH Baayen, K Postek - arXiv preprint arXiv:2007.07062, 2020 - arxiv.org
Non-convex optimal control problems occurring in, eg, water or power systems, typically
involve a large number of variables related through nonlinear equality constraints. The ideal …
involve a large number of variables related through nonlinear equality constraints. The ideal …