arXiv research

This work optimizes RL algorithms using entropy regularisation for continuous-time LQ problems.

Paper solves MV portfolio selection in jump-diffusion models with no-shorting constraint.

Study on PG learning for LQ MFC problems with common noise, proving convergence and sample complexity.

This paper develops q-learning methods for mean-field control problems.

Choquet regularization improves exploration in RL.

Paper proposes a RL approach for ALM with superior performance.

New theory extends LQ control to non-exponential discount scenarios.

This paper is devoted to study the effects arising from imposing a value-at-risk (VaR) constraint in mean-variance portfolio selection problem for an investor who receives a stochastic cash flow which he/she must then invest in a continuous-time financial market. For simplicity, we assume that there is only one investm…

This paper studies a class of continuous-time scalar-state stochastic Linear-Quadratic (LQ) optimal control problem with the linear control constraints. Applying the state separation theorem induced from its special structure, we develop the explicit solution for this class of problem. The revealed optimal control poli…

We solve a complex trade execution problem by simplifying it into a known LQ control problem.

RL solves discrete LQ control with Gaussian optimal policy.

Article presents QR and LQ decomposition algorithms for various matrix sizes and ranks.

Study of LQ MFGs in infinite-dimensional Hilbert spaces.

Developed LQ MFG theory with common noise, proving existence and uniqueness.

We study the global convergence of policy optimization for finding the Nash equilibria (NE) in zero-sum linear quadratic (LQ) games. To this end, we first investigate the landscape of LQ games, viewing it as a nonconvex-nonconcave saddle-point problem in the policy space. Specifically, we show that despite its nonconve…

We consider the exploration-exploitation tradeoff in linear quadratic (LQ) control problems, where the state dynamics is linear and the cost function is quadratic in states and controls. We analyze the regret of Thompson sampling (TS) (a.k.a. posterior-sampling for reinforcement learning) in the frequentist setting, i.…

Neural operators learn to solve LQ MFGs efficiently in infinite dimensions.

In this paper we introduce the notion of cofrontal mappings, as the dual objects to frontal mappings, and study their basic local and global properties. Cofrontals are very special mappings and far from generic nor stable except for the case of submersions. It is observed that any smooth mapping can be $C^0$-approximat…

We study the problem of adaptive control of a high dimensional linear quadratic (LQ) system. Previous work established the asymptotic convergence to an optimal controller for various adaptive control schemes. More recently, for the average cost LQ problem, a regret bound of ${O}(\sqrt{T})$ was shown, apart form logarit…

New algorithm reduces learning regret in multi-agent systems with unknown dynamics.

The paper solves a complex control problem with stochastic elements and switching conditions.

We study the performance of the certainty equivalent controller on Linear Quadratic (LQ) control problems with unknown transition dynamics. We show that for both the fully and partially observed settings, the sub-optimality gap between the cost incurred by playing the certainty equivalent controller on the true system …

We study in this paper a class of constrained linear-quadratic (LQ) optimal control problem formulations for the scalar-state stochastic system with multiplicative noise, which has various applications, especially in the financial risk management. The linear constraint on both the control and state variables considered…

We discuss two generalizations of the collar lemma. The first is the stable neighborhood theorem which says that a (not necessarily simple) closed geodesic in a hyperbolic surface has a \lq\lq stable neighborhood\rq\rq whose width only depends on the length of the geodesic. As an application, we show that there is a lo…

Model-free approaches for reinforcement learning (RL) and continuous control find policies based only on past states and rewards, without fitting a model of the system dynamics. They are appealing as they are general purpose and easy to implement; however, they also come with fewer theoretical guarantees than model-bas…

Over the moduli space of rank $n$ semi-stable lattices is a universal family of tori. Along the fibers, there are natural differential operators and differential equations, particularly, the heat equations and the Fokker-Planck equations in statistical mechanics. In this paper, we explain why, by taking averages over t…

New algorithms improve blind source separation for linear-quadratic mixtures.

A new method solves complex control problems with random coefficients.

Faster policy learning via continuous-time gradients.

Continuous time framework for discrete data denoising models.

Post-training optimizes model performance beyond base model limits.

Develops DPG methods for continuous-time RL with deterministic policies.

Study analyzes a new algorithm for complex optimization problems.

Motivated by the study of linear quadratic optimal control problems, we consider a dynamical system with a constant, quadratic Hamiltonian, and we characterize the number of conjugate times in terms of the spectrum of the Hamiltonian vector field $\vec{H}$. We prove the following dichotomy: the number of conjugate time…

Unitons, i.e.\ harmonic spheres in a unitary group, correspond to \lq uniton bundles\rq, i.e.\ holomorphic bundles over the compactified tangent space to the complex line with certain triviality and other properties. In this paper, we use a monad representation similar to Donaldson's representation of instanton bundles…

Continuous-time algorithms improve online learning performance.

This paper formulates and studies a stochastic maximum principle for forward-backward stochastic Volterra integral equations (FBSVIEs in short), while the control area is assumed to be convex. Then a linear quadratic (LQ in short) problem for backward stochastic Volterra integral equations (BSVIEs in short) is present …

Paper develops a continuous-time framework for financial markets without stochastic calculus.

Stochastic gradient descent in continuous time (SGDCT) provides a computationally efficient method for the statistical learning of continuous-time models, which are widely used in science, engineering, and finance. The SGDCT algorithm follows a (noisy) descent direction along a continuous stream of data. SGDCT performs…

New formula for 3-manifold invariants using combinatorial methods.

Consider power utility maximization of terminal wealth in a 1-dimensional continuous-time exponential Levy model with finite time horizon. We discretize the model by restricting portfolio adjustments to an equidistant discrete time grid. Under minimal assumptions we prove convergence of the optimal discrete-time strate…

Continuous-time MBRL framework tackles control systems with Bayesian ODEs.

Continuous-time model shows insider trading constraints impact market dynamics.

This paper examines the problem of learning with a finite and possibly large set of p base kernels. It presents a theoretical and empirical analysis of an approach addressing this problem based on ensembles of kernel predictors. This includes novel theoretical guarantees based on the Rademacher complexity of the corres…

Continuous-time analysis shows SGD with noise prefers flat minima.

New RL framework models continuous-time dynamics using neural ODEs.

Continuous-time SGD converges under certain conditions, useful for deep learning.

Logarithmic regret achieved in continuous-time linear-quadratic reinforcement learning.