arXiv research

Meta-algorithm optimizes nonstochastic bandits with infinitely many experts.

Algorithm detects changes online using expert tracking.

Bayesian algorithms improve online learning with adversaries over infinite action spaces.

A novel framework for regression with multiple experts, addressing challenges in infinite and continuous label spaces.

Unique solution found for even L^p Minkowski problem at p=p0, but fails for p<p0.

Paper proposes an efficient online learning method using an offline dataset for infinite horizon MDPs.

In mixtures-of-experts (ME) model, where a number of submodels (experts) are combined, there have been two longstanding problems: (i) how many experts should be chosen, given the size of the training data? (ii) given the total number of parameters, is it better to use a few very complex experts, or is it better to comb…

Expert augmentation improves hybrid model generalization.

The paper analyzes convergence rates of softmax gating in MoE models.

This note provides an easy construction of fake octagons.

We provide the first algorithm for online bandit linear optimization whose regret after T rounds is of order sqrt{Td ln N} on any finite class X of N actions in d dimensions, and of order d*sqrt{T} (up to log factors) when X is infinite. These bounds are not improvable in general. The basic idea utilizes tools from con…

We study a variant of decision-theoretic online learning in which the set of experts that are available to Learner can shrink over time. This is a restricted version of the well-studied sleeping experts problem, itself a generalization of the fundamental game of prediction with expert advice. Similar to many works in t…

Higher granularity in MoE models boosts expressivity exponentially.

The paper ranks experts based on task performance with noisy evaluations.

This work improves Gaussian process inference using mixtures of experts and nested SMC samplers.

In this work we study the problem of inferring a discrete probability distribution using both expert knowledge and empirical data. This is an important issue for many applications where the scarcity of data prevents a purely empirical approach. In this context, it is common to rely first on an initial domain knowledge …

Active learning (AL) is a learning paradigm where an active learner has to train a model (e.g., a classifier) which is in principal trained in a supervised way, but in AL it has to be done by means of a data set with initially unlabeled samples. To get labels for these samples, the active learner has to ask an oracle (…

This paper reports applications of Difference of Convex functions (DC) programming to Learning from Demonstrations (LfD) and Reinforcement Learning (RL) with expert data. This is made possible because the norm of the Optimal Bellman Residual (OBR), which is at the heart of many RL and LfD algorithms, is DC. Improvement…

The study finds infinitely many different geometries for odd-dimensional manifolds with positive Ricci curvature.

New algorithm minimizes expert selection regret in partial bandit feedback.

Paper studies continuous prediction with experts' advice using differential equations.

This paper investigates optimal trading strategies in a financial market with multidimensional stock returns where the drift is an unobservable multivariate Ornstein-Uhlenbeck process. Information about the drift is obtained by observing stock returns and expert opinions. The latter provide unbiased estimates on the cu…

A new method for learning to defer decisions with expert advice improves over standard methods.

Many predictive tasks, such as diagnosing a patient based on their medical chart, are ultimately defined by the decisions of human experts. Unfortunately, encoding experts' knowledge is often time consuming and expensive. We propose a simple way to use fuzzy and informal knowledge from experts to guide discovery of int…

EBBS integrates expert assessments into MIO best-subsets problem.

Improved transfer learning with expert models, reducing compute and speeding up performance.

New technique reduces imitation learning performance gap in finite samples.

Accessible groups with infinitely many ends have infinitely many twisted conjugacy classes.

AceIRL learns expert reward from active exploration.

Generalized algorithm for translation and scale-invariant prediction.

LEAQI learns to query an expert less often for noisy guidance.

For the prediction with experts' advice setting, we construct forecasting algorithms that suffer loss not much more than any expert in the pool. In contrast to the standard approach, we investigate the case of long-term forecasting of time series and consider two scenarios. In the first one, at each step $t$ the learne…

We propose an expert-augmented actor-critic algorithm, which we evaluate on two environments with sparse rewards: Montezumas Revenge and a demanding maze from the ViZDoom suite. In the case of Montezumas Revenge, an agent trained with our method achieves very good results consistently scoring above 27,000 points (in ma…

We revisit the issue of the existence of infinitely many distinct prime knots with the same Alexander invariant. We present infinitely many distinct families, each family made up of infinitely many distinct knots. Within each family, the Alexander invariant is the same. Unlike other examples in the literature, ours are…

Veech groups uniformize Teichmüller geodesic curves in Riemann moduli space. Recently, examples of infinitely generated Veech groups have been given. We show that these can even have infinitely many cusps and infinitely many infinite ends. We further show that examples exist for which each direction of an infinite end …

AGFN improves causal discovery by integrating expert feedback and handling latent confounding.

A new method for sparse Gaussian process regression using correlated experts.

The paper introduces a method to incorporate expert opinion on observable quantities into statistical models.

Positive braids have endless filling possibilities.

Novel algorithms for online learning with uncertain feedback graphs reduce regret.

CoCoAFusE fuses expert predictions to model complex patterns with interpretability and uncertainty.

The author, in her previous paper, constructed an infinite family of 3-bridge links each of which admits infinitely many 3-bridge spheres up to isotopy. In this paper, we prove that if a prime, unsplittable link $L$ in $S^3$ admits infinitely many 3-bridge spheres up to isotopy then $L$ belongs to the family.

We prove that all maximal-tb Legendrian torus links (n,m) in the standard contact 3-sphere, except for (2,m),(3,3),(3,4) and (3,5), admit infinitely many Lagrangian fillings in the standard symplectic 4-ball. This is proven by constructing infinite order Lagrangian concordances which induce faithful actions of the modu…

3-manifolds have covers with infinitely many ideal triangulations.

In this article we present infinitely many 3-manifolds admitting infinitely many universally tight contact structures each with trivial Ozsvath-Szabo contact invariants. By known properties of these invariants the contact structures constructed here are non weakly symplectically fillable.

New examples of Legendrian links with infinitely many fillings.

A manifold has infinitely many sphere fibrations over a sphere.

Consider a regression problem where there is no labeled data and the only observations are the predictions $f_i(x_j)$ of $m$ experts $f_{i}$ over many samples $x_j$. With no knowledge on the accuracy of the experts, is it still possible to accurately estimate the unknown responses $y_{j}$? Can one still detect the leas…