Logarithmic connections on principal bundles over normal varieties are studied.
problem Existence and properties of logarithmic connections on principal bundles over normal varieties.
method Introducing logarithmic connections, showing equivalence to covariant derivatives, and proving existence conditions.
result Existence of logarithmic connections on principal bundles over normal varieties is equivalent to certain conditions on the associated vector bundles and adjoint bundles.
Paper proposes FedQ-Advantage for federated Q-learning with near-optimal regret and low communication cost.
problem Near-optimal federated Q-learning with low communication cost.
method Reference-advantage decomposition for variance reduction, synchronization between agents and server, policy update.
result Achieves almost optimal regret and near-linear regret speedup compared to single-agent learning.
Study quadratic one-forms on logarithmic Higgs bundles on pointed curves.
problem Quadratic one-forms on logarithmic Higgs bundles on pointed curves.
method Use elementary pole cancellation for invariant polynomials.
result Found a logarithmic quadratic one-form.
Study logarithmic flat connections on principal bundles using Lie groupoids.
problem Classify flat connections on principal bundles with logarithmic singularities.
method Use tools from Lie groupoid theory to classify representations and establish van Kampen theorems.
result Obtain a functorial Riemann-Hilbert correspondence for logarithmic connections.
We create a new online reduction of multiclass classification to binary classification for which training and prediction time scale logarithmically with the number of classes. Compared to previous approaches, we obtain substantially better statistical performance for two reasons: First, we prove a tighter and more comp…
Normal forms and moduli stacks for flat connections on complex manifolds.
problem Understanding singular flat connections on complex manifolds.
method Introducing homogeneous Lie groupoids and studying their representation theory to prove normal form theorems and moduli space structures.
result Moduli spaces of singular flat connections admit the structure of algebraic quotient stacks.
The paper proves a new version of dimensional reduction in cohomological Donaldson-Thomas theory.
problem Proving a new version of dimensional reduction in cohomological Donaldson-Thomas theory.
method Using cohomological Donaldson-Thomas theory and loop stacks of 0-shifted symplectic stacks.
result Shows the BPS cohomology of loop stacks admits a description analogous to orbifold cohomology.
Proves a Baum--Bott formula for foliations by curves with logarithmic terms.
problem Analyzing singularities and smoothness in foliations by curves.
method Logarithmic Baum--Bott residues for foliated triples (X,F,D), relating to Poincaré's Problem and GSV indices. result Logarithmic Baum--Bott residues generalize Aleksandrov logarithmic index for vector fields on hypersurfaces.
Greedy pruning reduces neural networks by a logarithmic number of tickets, improving accuracy.
problem Pruning large neural networks to reduce size while maintaining accuracy.
method Greedy optimization-based pruning method with exponential decay guarantee.
result The discrepancy between pruned and original networks decays exponentially with network size.
This paper is focused on geometric aspects of two particular types of finite-variable reductions in the dispersionless Toda hierarchy. The reductions are formulated in terms of "Landau-Ginzburg potentials" that play the role of reduced Lax functions. One of them is a generalization of Dubrovin and Zhang's trigonometric…
Efficient algorithm for contextual bandits with first-order guarantees.
problem Adapting to low noise in contextual bandits.
method Reduction to online regression with logarithmic loss.
result Optimal and efficient first-order guarantees for contextual bandits.
We consider saddle point problems which objective functions are the average of n strongly convex-concave individual components. Recently, researchers exploit variance reduction methods to solve such problems and achieve linear-convergence guarantees. However, these methods have a slow convergence when the condition n…
Study on optimal rates for sequential probability assignment using smoothed analysis.
problem Optimal rates for sequential probability assignment under smoothed adversaries.
method General-purpose reduction from minimax rates to transductive learning, development of an efficient algorithm using MLE oracle.
result Optimal (logarithmic) fast rates for parametric and finite VC dimension classes, sublinear regret for general classes.
Analytic torsion behavior studied for degenerating manifolds with equivariant bundles.
problem Behavior of analytic torsion for degenerating manifolds with equivariant bundles.
method Asymptotic expansion of equivariant analytic torsion, Quillen metrics, L2-metrics, Bott-Chern classes.
result Leading term of analytic torsion has logarithmic singularity, subdominant term has loglog-type singularity.
Bandits with Knapsacks (BwK) is a general model for multi-armed bandits under supply/budget constraints. While worst-case regret bounds for BwK are well-understood, we present three results that go beyond the worst-case perspective. First, we provide upper and lower bounds which amount to a full characterization for lo…
Study sparsity benefits in infinite feature contextual bandits.
problem Minimizing regret in infinite feature contextual bandits.
method Novel reduction to multi-armed bandits, Feel-Good Thompson Sampling algorithm.
result Regret bounds match lower bounds up to logarithmic factors, logarithmic dependence on effective features.
Method identifies low-dimensional structure in high-dimensional probability measures.
problem Identifying low-dimensional structure in high-dimensional probability measures.
method Extends prior work on minimizing majorizations of the Kullback-Leibler divergence to identify optimal approximations within a specific class of measures.
result Connection between dimensional logarithmic Sobolev inequality and approximations with the ansatz.
New algorithm reduces offline RL data requirements significantly.
problem Optimizing policies using only historical data in reinforcement learning.
method Off-Policy Double Variance Reduction (OPDVR) algorithm.
result OPDVR achieves optimal sample complexity with O(H2/dmε2) episodes. We discuss the `hd-compactification' of a semi-simple Lie group to a manifold with corners; it is the real analog of the wonderful compactification of deConcini and Procesi. There is a 1-1 correspondence between the boundary faces of the compactification and conjugacy classes of parabolic subgroups with the boundary fa…
Develops a new geometric framework for quantum metrics.
problem Quantum metric generalization for pure two-qubit states.
method Support-projected Petz monotone geometry for pure two-qubit families.
result Strictly generalizes SLD/Bures case and includes other metrics.
We consider the problem of estimating the conditional probability of a label in time O(log n), where n is the number of possible labels. We analyze a natural reduction of this problem to a set of binary regression problems organized in a tree structure, proving a regret bound that scales with the depth of the tree. Mot…
A local Riemann-Hilbert correspondence for tame meromorphic connections on a curve compatible with a parahoric level structure will be established. Special cases include logarithmic connections on G-bundles and on parabolic G-bundles, where G is a complex reductive group. The corresponding Betti data involves pairs (M,…
Paper proposes an algorithm to optimize CVaR using retrospective approximation and importance sampling.
problem Optimizing risk-averse problems with large sample requirements for CVaR.
method Retrospective approximation combined with importance sampling, tailored for CVaR optimization.
result The proposed algorithm reduces variance efficiently and is computationally efficient.
The paper analyzes portfolio credit risk using Archimedean copulas and introduces efficient simulation methods.
problem Analyzing large losses from credit portfolio defaults with Archimedean copulas.
method Derives asymptotic results and develops variance reduction algorithms for Monte Carlo simulations.
result Proposed algorithms significantly enhance classical Monte Carlo methods for estimating portfolio credit risk.
Causal discovery from empirical data is a fundamental problem in many scientific domains. Observational data allows for identifiability only up to Markov equivalence class. In this paper we first propose a polynomial time algorithm for learning the exact correctly-oriented structure of the transitive reduction of any c…
Stochastic convex optimization algorithms are the most popular way to train machine learning models on large-scale data. Scaling up the training process of these models is crucial, but the most popular algorithm, Stochastic Gradient Descent (SGD), is a serial method that is surprisingly hard to parallelize. In this pap…
In this paper, we introduce the notions of logarithmic Poisson structure and logarithmic principal Poisson structure; we prove that the latter induces a representation by logarithmic derivation of the module of logarithmic Kahler differentials; therefore, it induces a differential complex from which we derive the notio…
New approach achieves optimal rates for differentially private stochastic convex optimization with heavy-tailed gradients.
problem Differentially private stochastic convex optimization with heavy-tailed gradients.
method Reduction-based approach to achieve optimal rates.
result Achieved optimal rates up to logarithmic factors, nearly matching a lower bound.
In their papers published in 1993 and 1994, by expressing certain physical quantity in two distinct ways, Bershadsky-Cecotti-Ooguri-Vafa discovered a remarkable equivalence between Ray-Singer analytic torsion and elliptic instanton numbers for Calabi-Yau threefolds. After their discovery, in a paper published in 2008, …
Study on Higgs bundles and hyperpolygon spaces using Hitchin metrics.
problem Investigating the Hitchin metric on moduli spaces of Higgs bundles.
method Using Hitchin hyperkähler metric and parabolic Deligne-Hitchin moduli space.
result Rescaled Hitchin metric converges to hyperpolygon space's hyperkähler metric in the semiclassical limit.
It is now known that an extended Gaussian process model equipped with rescaling can adapt to different smoothness levels of a function valued parameter in many nonparametric Bayesian analyses, offering a posterior convergence rate that is optimal (up to logarithmic factors) for the smoothness class the true function be…
Study on geodesic distances on SE(3)/SO(2) in machine learning.
problem Investigating the efficiency of computationally efficient sections in selecting geodesic distances.
method Analyzing geodesic distances on reductive homogeneous spaces, proving the efficiency of minimal distance sections.
result Minimal distance sections are not always geodesic minimizers, but minimal horizontal geodesics are.
We consider the problem of efficient randomized dimensionality reduction with norm-preservation guarantees. Specifically we prove data-dependent Johnson-Lindenstrauss-type geometry preservation guarantees for Ho's random subspace method: When data satisfy a mild regularity condition -- the extent of which can be estima…
Study rigidity by logarithmic capacity and related functions.
problem Rigidity phenomena in kernel functions and capacities.
method Exploration of Bergman kernel, logarithmic capacity, Green's function, and Euclidean distance/volume.
result Established rigidity theorems by logarithmic capacity.
New theory shows how multi-head attention reduces variance and decorrelates outputs.
problem Understanding and optimizing multi-head attention in neural networks.
method Developed a statistical theory linking multi-head attention to ensemble Nadaraya-Watson estimators.
result MHA variance reduction depends on head decorrelation, not just head count.
Study online linear regression with paid noise reduction.
problem Online linear regression with noisy features and the ability to pay for reduced noise.
method Analyzes regret against optimal predictor, uses matrix martingale concentration.
result Optimal regret rates for known and unknown noise covariance.
We adopt data structure in the form of cover trees and iteratively apply approximate nearest neighbour (ANN) searches for fast compressed sensing reconstruction of signals living on discrete smooth manifolds. Levering on the recent stability results for the inexact Iterative Projected Gradient (IPG) algorithm and by us…
Study real logarithms of semi-simple matrices, focusing on differential structure.
problem Understanding the differential structure of real logarithms of semi-simple matrices.
method Examines the differential structure of real logarithms of semi-simple matrices under specific matrix types.
result Characterizes the differential structure of real logarithms of semi-simple matrices.
Study excess logarithmic residues for foliations to bound invariant hypersurfaces and test log canonicity.
problem Bounding invariant hypersurfaces and testing log canonicity of singularities.
method Introduce excess logarithmic residues, prove residue formula, derive Poincaré-type bound, and use them to recover log discrepancies.
result Componentwise logarithmic residues of a lifted foliation along the exceptional divisor recover log discrepancies of singularities.
We present a new method to solve certain ∂ˉ-equations for logarithmic differential forms by using harmonic integral theory for currents on Kahler manifolds. The result can be considered as a ∂ˉ-lemma for logarithmic forms. As applications, we generalize the result of Deligne about closedness…
Local logarithmic Brunn-Minkowski holds for zonoids.
problem Logarithmic Brunn-Minkowski conjecture for zonoids
method Bochner method variant
result Local form of conjecture proven for zonoids
The paper proposes differentially private sliced inverse regression algorithms for high-dimensional data.
problem Privacy concerns in high-dimensional data analysis.
method Differentially private sliced inverse regression algorithms designed for privacy preservation.
result Achieves minimax lower bounds up to logarithmic factors.
New framework for logarithmically divergent integrals on manifolds with corners.
problem Logarithmically divergent integrals on manifolds with corners.
method Introduces new geometric framework and morphisms in logarithmic geometry.
result Functorial characterization of regularized integration.
New RL algorithms reduce costs for single-agent and federated learning.
problem Minimizing costs in RL and federated RL settings.
method Q-EarlySettled-LowCost and FedQ-EarlySettled-LowCost algorithms.
result First algorithms to achieve low burn-in and logarithmic switching costs.
Directly proves logarithmic systolic growth for all hyperbolic surfaces.
problem Proving logarithmic systolic growth for all hyperbolic surfaces.
method Using original Brooks/Buser-Sarnak surfaces through a direct approach.
result Directly proves logarithmic systolic growth for all hyperbolic surfaces.
Polynomial-time algorithm estimates edge density of random graphs with privacy and robustness.
problem Estimating edge density of random graphs while maintaining privacy and robustness.
method Sum-of-squares algorithm for robust edge density estimation and reduction from privacy to robustness.
result Optimal error rate up to logarithmic factors, matching theoretical lower bounds.
We design and study a Contextual Memory Tree (CMT), a learning memory controller that inserts new memories into an experience store of unbounded size. It is designed to efficiently query for memories from that store, supporting logarithmic time insertion and retrieval operations. Hence CMT can be integrated into existi…
Logarithmic separation profile in hyperbolic groups shows hierarchical structure.
problem Understanding hierarchical structure in hyperbolic groups with logarithmic separation.
method Proving groups with logarithmic separation split over cyclic groups and providing counterexamples.
result Not all groups with hierarchical structure have logarithmic separation profile.