arXiv research

New RL method learns to skip states in linearly $q^π$-realizable MDPs, simplifying to linear MDPs.

RNN-HAR model improves VaR forecasting with long-memory and non-linear dynamics.

Paper tackles sample-efficient RL for linearly realizable MDPs with limited revisiting.

TensorPlan algorithm finds δ-optimal policies with poly$(H,d)$ queries under linearly realizable state-value function.

Efficient RL algorithm for MDPs with linear $Q^π$ realizability, achieving optimal regret bound.

New superbridge index calculations for knots with odd edges.

This article presents an improvement and extension of the heuristic first presented by Hougardy, Lutz, and Zelke in 2010 for realizing triangulated orientable surfaces with few vertices by a simplex-wise linear embedding. The improvement consists in the applicability to non-orientable surfaces (simplex-wise linear imme…

We study distributions of realized variance (squared realized volatility) and squared implied volatility, as represented by VIX and VXO indices. We find that Generalized Beta distribution provide the best fits. These fits are much more accurate for realized variance than for squared VIX and VXO -- possibly another indi…

New bandit algorithm works without realizability assumption.

New method efficiently evaluates policies using trajectory data.

Study bounds noise level in linear regression with dependent data.

Paper tackles MLR prediction error without assuming realizable models.

Smooth deformation space of Coxeter polytopes proven for orderable orbifolds.

We introduce a class of k-potential submanifolds in pseudo-Euclidean spaces and prove that for an arbitrary positive integer k and an arbitrary nonnegative integer p, each N-dimensional Frobenius manifold can always be locally realized as an N-dimensional k-potential submanifold in ((k + 1) N + p)-dimensional pseudo-Eu…

We consider the problem of learning a realization for a linear time-invariant (LTI) dynamical system from input/output data. Given a single input/output trajectory, we provide finite time analysis for learning the system's Markov parameters, from which a balanced realization is obtained using the classical Ho-Kalman al…

In this work, we highlight a connection between the incremental proximal method and stochastic filters. We begin by showing that the proximal operators coincide, and hence can be realized with, Bayes updates. We give the explicit form of the updates for the linear regression problem and show that there is a one-to-one …

Bayesian model improves asset price forecasting using realized volatility.

The paper provides a combinatorial criterion for realizing tropical pluri-canonical divisors.

Motivated by the importance and universal character of phase singularities which are clarified recently, we study the local structure of equi-phase loci near the dislocation locus of complex valued planar and spatial waves, from the viewpoint of singularity theory of differentiable mappings, initiated by H. Whitney and…

We show that each classical pseudoriemann symmetric space G/H can be realized as space of pairs of complementary subspaces in a linear space. For each classical symmetric space we construct an open embedding to a grassmannian or to a product of two grassmanianns. We also show that the representation of the group G in L…

Trajectory data suffices for efficient RL in linear MDPs.

Geometric framework for signed multivariate tail-dependence compatibility at various thresholds.

Constructs equivariant embeddings of Hermitian symmetric spaces into tangent spaces.

This study connects ReLU neural networks to toric geometry to analyze function realization.

This paper develops a general method for constructing Poisson integrators.

Optimal algorithm for maximizing rewards in contextual bandits with resource constraints.

Efficient algorithm for near-optimal online learning with generalized linear functions.

Study of loss functions for learning to defer, proving consistency.

Paper proposes a deep learning method for better covariance matrix forecasting.

This paper investigates how the conditional quantiles of future returns and volatility of financial assets vary with various measures of ex-post variation in asset prices as well as option-implied volatility. We work in the flexible quantile regression framework and rely on recently developed model-free measures of int…

New algorithm tackles non-linear utility in MNL bandits with $ilde{O}(\sqrt{T})$ regret.

New algorithm trains ReLU gates provably in linear time.

In this paper, we give a weak classification of locally linear pseudofree actions of the cyclic group of order 3 on a $K3$ surface, and prove the existence of such an action which can not be realized as a smooth action on the standard smooth $K3$ surface.

New efficient method for inverse Z-transform reduces complexity significantly.

New algorithm SELECT minimizes satisficing regret in bandits.

Graph manifolds' Thurston norms are sums of linear functionals, and every such norm can be realized.

The study classifies and characterizes totally symmetric sets in the general linear group.

Characterizes inductive bias in multi-channel linear CNNs with bounded weight norm.

New method tests independence with single nonstationary time series.

The problem of immersing a simply connected surface with a prescribed shape operator is discussed. From classical and more recent work, it is known that, aside from some special degenerate cases, such as when the shape operator can be realized by a surface with one family of principal curves being geodesic, the space o…

New framework tackles stochastic latent subgroup heterogeneity in online decision-making.

Analyzes multi-day stock returns, showing linear volatility and mean dependence.

We present a geometric construction of central extensions of covering groups of the group of volume preserving diffeomorphisms, integrating central extensions of the Lie algebra of divergence free vector fields defined by Lichnerowicz cocycles. Certain covering spaces of non-linear Grassmannians can be realized as preq…

HAR model outperforms ML in stock forecasting with correct fitting schemes.

Oracle-efficient algorithm for offline RL with partial data coverage.

TensorPlan shows an exponential lower bound for planning in MDPs with linearly realizable value functions.

The compact simply connected Riemannian 4-symmetric spaces were classified by J.A. Jim{é}nez. As homogeneous manifolds, these spaces are of the $G/H$, where $G$ is a connected compact simple Lie group with an automorphism $\tildeγ$ of oder 4 and $H$ is a fixed points subgroup $G^γ$ of $G$. In the present article, for t…

This paper presents the nonparametric inference for nonlinear volatility functionals of general multivariate Itô semimartingales, in high-frequency and noisy setting. Pre-averaging and truncation enable simultaneous handling of noise and jumps. Second-order expansion reveals explicit biases and a pathway to bias correc…