arXiv research

New algorithm solves utility maximization with deep learning for constrained problems.

In this paper we study mean-field type control problems with risk-sensitive performance functionals. We establish a stochastic maximum principle (SMP) for optimal control of stochastic differential equations (SDEs) of mean-field type, in which the drift and the diffusion coefficients as well as the performance function…

This work presents the results of an empirical research with the target of modeling the stylized facts of the daily expost System Marginal Price (SMP) of the Greek wholesale electricity market, using data from January 2004 to December of 2011. SMP is considered here as the footprint of an underline stochastic and nonli…

SMP estimator improves density and logistic regression under misspecification.

Paper proves deep learning method for stochastic control converges and outperforms existing algorithms.

Score-based martingale posteriors improve uncertainty quantification in deep neural networks.

SMP model preserves proximity and permutation in graph neural networks.

We establish a stochastic maximum principle (SMP) for control problems of partially observed diffusions of mean-field type with risk-sensitive performance functionals.

Mathematical model audits social media algorithms to prevent bias.

Matrix Product States (MPS), also known as Tensor Train (TT) decomposition in mathematics, has been proposed originally for describing an (especially one-dimensional) quantum system, and recently has found applications in various applications such as compressing high-dimensional data, supervised kernel linear classifie…

Let $M$ be a connected orientable compact surface, $f:M\to\mathbb{R}$ be a Morse function, and $\mathcal{D}_{\mathrm{id}}(M)$ be the group of difeomorphisms of $M$ isotopic to the identity. Denote by $\mathcal{S}'(f)=\{f\circ h = f\mid h\in\mathcal{D}_{\mathrm{id}}(M)\}$ the subgroup of $\mathcal{D}_{\mathrm{id}}(M)$ c…

Utilization of non-linear tools to characterize the state of development of the electricity markets in Italy and Greece. This is equivalent to testing the Efficient Market Hypothesis on these markets. The tools include a variety of complexity measures like Maximal Lyapunov and Hurst exponents and HHI index for market c…

DanSmp predicts stock movement using a hybrid-relational MKG and dual attention networks.

Unified policy controls diverse agents through modular neural networks.

NDDV estimates data point value from a single stochastic trajectory.