Saved in:
Bibliographic Details
Main Authors: Ludkovski, Michael, Xie, Changgen, Zhu, Zimu
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.04309
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915658644586496
author Ludkovski, Michael
Xie, Changgen
Zhu, Zimu
author_facet Ludkovski, Michael
Xie, Changgen
Zhu, Zimu
contents We consider numerical resolution of principal-agent (PA) problems in continuous time. We formulate a generic PA model with continuous and lump payments and a multi-dimensional strategy of the agent. To tackle the resulting Hamilton-Jacobi-Bellman equation with an implicit Hamiltonian we develop a novel deep learning method: the Deep Principal-Agent Actor Critic (DeepPAAC) Actor-Critic algorithm. DeepPAAC is able to handle multi-dimensional states and controls, as well as constraints. We investigate the role of the neural network architecture, training designs, loss functions, etc. on the convergence of the solver, presenting five different case studies.
format Preprint
id arxiv_https___arxiv_org_abs_2511_04309
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DeepPAAC: A New Deep Galerkin Method for Principal-Agent Problems
Ludkovski, Michael
Xie, Changgen
Zhu, Zimu
Numerical Analysis
Machine Learning
91B43, 68T07, 93E20
We consider numerical resolution of principal-agent (PA) problems in continuous time. We formulate a generic PA model with continuous and lump payments and a multi-dimensional strategy of the agent. To tackle the resulting Hamilton-Jacobi-Bellman equation with an implicit Hamiltonian we develop a novel deep learning method: the Deep Principal-Agent Actor Critic (DeepPAAC) Actor-Critic algorithm. DeepPAAC is able to handle multi-dimensional states and controls, as well as constraints. We investigate the role of the neural network architecture, training designs, loss functions, etc. on the convergence of the solver, presenting five different case studies.
title DeepPAAC: A New Deep Galerkin Method for Principal-Agent Problems
topic Numerical Analysis
Machine Learning
91B43, 68T07, 93E20
url https://arxiv.org/abs/2511.04309