arXiv research

New algorithm POO optimizes noisy, unknown-smooth functions.

Kernel smoothing on unknown manifolds with bounds and asymptotic normality.

Optimal nonparametric regression estimator adapts to unknown smoothness.

We tackle tensor denoising with unknown permutations, achieving optimal recovery with polynomial estimators.

We study a non-parametric multi-armed bandit problem with stochastic covariates, where a key complexity driver is the smoothness of payoff functions with respect to covariates. Previous studies have focused on deriving minimax-optimal algorithms in cases where it is a priori known how smooth the payoff functions are. I…

SATL adapts to varying smoothness in hypothesis transfer learning.

Spike-and-Slab Deep Learning (SS-DL) is a fully Bayesian alternative to Dropout for improving generalizability of deep ReLU networks. This new type of regularization enables provable recovery of smooth input-output maps with unknown levels of smoothness. Indeed, we show that the posterior distribution concentrates at t…

This paper, to be regularly updated, lists those prime knots with the fewest possible number of crossings for which values of basic knot invariants, such as the unknotting number or the smooth 4-genus, are unknown. This list is being developed in conjunction with "KnotInfo" (www.indiana.edu/~knotinfo), a web-based tabl…

This paper contains the results of efforts to determine values of the smooth and the topological slice genus of 11- and 12-crossing knots. Upper bounds for these genera were produced by using a computer to search for genus one concordances between knots. For the topological slice genus further upper bounds were produce…

New algorithm adapts to unknown demand smoothness for dynamic pricing.

Computed the 4-genus for all 12-crossing prime knots.

New method infers causal effects without knowing control variables.

In the present paper, we studied a Dynamic Stochastic Block Model (DSBM) under the assumptions that the connection probabilities, as functions of time, are smooth and that at most $s$ nodes can switch their class memberships between two consecutive time points. We estimate the edge probability tensor by a kernel-type p…

By studying the Heegaard Floer homology of the preimage of a knot K in S^3 inside its double branched cover, we develop simple obstructions to K having finite order in the classical smooth concordance group. As an application, we prove that all 2-bridge knots of crossing number at most 12 for which the smooth concordan…

Method estimates treatment effects in dyadic data with unknown confounders.

New algorithm reduces prediction error in online learning without knowing base measure.

Interesting theoretical associations have been established by recent papers between the fields of active learning and stochastic convex optimization due to the common role of feedback in sequential querying mechanisms. In this paper, we continue this thread in two parts by exploiting these relations for the first time …

Smoothing splines provide a powerful and flexible means for nonparametric estimation and inference. With a cubic time complexity, fitting smoothing spline models to large data is computationally prohibitive. In this paper, we use the theoretical optimal eigenspace to derive a low rank approximation of the smoothing spl…

We use the Heegaard Floer obstructions defined by Grigsby, Ruberman, and Strle to show that forty-six of the sixty-seven knots through eleven crossings whose concordance orders were previously unknown have infinite concordance order.

New method for adaptive estimation and inference in econometric models without knowing smoothness.

We study contextual bandit learning with an abstract policy class and continuous action space. We obtain two qualitatively different regret bounds: one competes with a smoothed version of the policy class under no continuity assumptions, while the other requires standard Lipschitz assumptions. Both bounds exhibit data-…

New matching estimators correct bias in multivariate settings without smoothing parameters.

Consider a nonparametric contextual multi-arm bandit problem where each arm $a \in [K]$ is associated to a nonparametric reward function $f_a: [0,1] \to \mathbb{R}$ mapping from contexts to the expected reward. Suppose that there is a large set of arms, yet there is a simple but unknown structure amongst the arm reward…

A boosting method improves nonparametric density estimation without smoothing assumptions.

Let $\mathbf{x}_j = \mathbfθ + \mathbfε_j$, $j=1,\dots,n$ be i.i.d. copies of a Gaussian random vector $\mathbf{x}\sim\mathcal{N}(\mathbfθ,\mathbfΣ)$ with unknown mean $\mathbfθ \in \mathbb{R}^d$ and unknown covariance matrix $\mathbfΣ\in \mathbb{R}^{d\times d}$. The goal of this article is to study the estimation of $…

Adaptive data fusion boosts efficiency in multi-task optimization.

Improved SGD with AdaGrad stepsizes adapts to unknown parameters and unbounded gradients.

Study approximates unknown function levels with queries.

We consider the problem of global optimization of an unknown non-convex smooth function with zeroth-order feedback. In this setup, an algorithm is allowed to adaptively query the underlying function at different locations and receives noisy evaluations of function values at the queried points (i.e. the algorithm has ac…

Modified ReLU networks improve regression estimation rates.

It is shown that bootstrap approximations of support vector machines (SVMs) based on a general convex and smooth loss function and on a general kernel are consistent. This result is useful to approximate the unknown finite sample distribution of SVMs by the bootstrap approach.

Improved mean estimation for symmetric distributions with finite-sample guarantees.

Paper introduces privacy-preserving inventory policy learning for feature-based newsvendor with unknown demand.

Smoothed analysis shows that many classes become learnable from positive-only samples.

New study reveals a polynomial penalty for adapting to unknown margin parameters in batched nonparametric bandits.

Let M be a smooth compact oriented manifold without boundary, imbedded in a euclidean space E and let f be a smooth map of M into a Riemannian manifold N. An unknown state x in M is observed via X=x+su where s>0 is a small parameter and u is a white Gaussian noise. For a given smooth prior on M and smooth estimators g …

SPH-ParVI uses fluid dynamics to sample unknown densities efficiently.

Bayesian methods estimate regression functions on submanifolds using graph Laplacian eigenbasis.

Study optimal control in unknown nonlinear systems with near-optimal regret bound.

Study shows overparameterization helps in generalizing from smooth interpolants.

New method for robust fixed-point smoothing without state augmentation.

We propose a principled algorithm for robust Bayesian filtering and smoothing in nonlinear stochastic dynamic systems when both the transition function and the measurement function are described by non-parametric Gaussian process (GP) models. GPs are gaining increasing importance in signal processing, machine learning,…

A graph manifold rational homology $3$-sphere $W$ with a left-orderable fundamental group admits a co-oriented taut foliation, though it is unknown whether it admits a smooth co-oriented taut foliation. In this paper we extend the gluing theorem of arXiv:1401.7726 to graph manifold rational homology solid tori and use …

A novel GPUM constructs Gaussian Processes for unknown manifolds with probabilistic metrics.

We use refined spectral sequence arguments to calculate known and previously unknown bi-Hamiltonian cohomology groups, which govern the deformation theory of semi-simple bi-Hamiltonian pencils of hydrodynamic type with one independent and \( N\) dependent variables. In particular, we rederive the result of Dubrovin-Liu…

The purpose of this work is to develop and study a distributed strategy for Pareto optimization of an aggregate cost consisting of regularized risks. Each risk is modeled as the expectation of some loss function with unknown probability distribution while the regularizers are assumed deterministic, but are not required…

ASE safely explores unknown MDPs with unknown dynamics, improving sample efficiency.

In this article, a three-time levels compact scheme is proposed to solve the partial integro-differential equation governing the option prices under jump-diffusion models. In the proposed compact scheme, the second derivative approximation of unknowns is approximated by the value of unknowns and their first derivative …