Lower bounds for higher-order methods in non-convex optimization.
problem Proving lower bounds for higher-order methods in smooth non-convex finite-sum optimization.
method Analyzing deterministic and randomized algorithms, proposing a new smoothness assumption.
result Proves optimal lower bounds for simulating pth-order regularized methods on the whole function.
New lower bounds for bilevel optimization with first-order oracles.
problem Complexity of bilevel optimization with first-order oracles.
method Development of hard instances and proof of lower bounds.
result Nontrivial lower bounds for first-order zero-respecting algorithms.
Paper improves stochastic bilevel optimization methods for highly-smooth problems.
problem Finding ε ε ε -stationary points in stochastic bilevel optimization. method Proposes F 2 {}^2 2 SA- p p p methods using p p p th-order finite differences for hyper-gradient approximation. result Achieves upper complexity bound of i l d e O ( p ε − 4 − p / 2 ) ilde{\mathcal{O}}(p ε^{-4-p/2}) i l d e O ( p ε − 4 − p /2 ) for p p p th-order smooth problems. New algorithm finds approximate stationary points in non-convex optimization.
problem Finding approximate stationary points in non-convex stochastic optimization.
method Design of an algorithm using O ( ε − 3 ) O(ε^{-3}) O ( ε − 3 ) stochastic gradient and Hessian-vector products. result Optimal rate of O ( ε − 3 ) O(ε^{-3}) O ( ε − 3 ) for finding ε ε ε -approximate stationary points, matching lower bounds. New algorithm optimizes convex functions with noisy evaluations in one dimension.
problem Optimizing convex functions with noisy zero-order evaluations in one dimension.
method Proposed a computationally efficient algorithm achieving O ( 1 / T ) O(1/\sqrt{T}) O ( 1/ T ) convergence rate. result Achieved the optimal O ( 1 / T ) O(1/\sqrt{T}) O ( 1/ T ) convergence rate, closing the gap in one dimension. Study noise-free kernel bandits, finding upper bounds on regret.
problem Optimizing unknown functions without noise.
method Upper bounds on regret for noise-free kernel-based bandits.
result No order optimal regret bounds are established, conjecture on optimal bound.
New algorithms optimize convex functions with high-order derivatives.
problem Optimizing convex functions with high-order derivatives under various norms.
method Developed a non-Euclidean inexact accelerated proximal point method using an inexact uniformly convex regularizer.
result Showed nearly optimal algorithms for high dimensions in the black-box oracle model for ℓ p \ell_p ℓ p -settings and all q ≥ 1 q \geq 1 q ≥ 1 . New algorithms reduce regret in online MDPs by adapting to data and variance.
problem Adapting to both adversarial and stochastic environments in online MDPs.
method Develops algorithms based on global optimization and policy optimization, using optimistic follow-the-regularized-leader with log-barrier regularization.
result Achieves refined data-dependent and variance-dependent regret bounds.
New algorithm for identifying optimal arms in stochastic bandit problems.
problem Optimal arm identification in stochastic bandit problems with many arms.
method Characterized optimal learning rates and provided algorithms with matching bounds.
result Lower bounds and matching upper bounds for cumulative regret and best-arm identification.
State-of-the-art methods in convex and non-convex optimization employ higher-order derivative information, either implicitly or explicitly. We explore the limitations of higher-order optimization and prove that even for convex optimization, a polynomial dependence on the approximation guarantee and higher-order smoothn…
New methods solve optimization problems with heavy-tailed noise, improving upon existing complexity bounds.
problem Optimization problems with heavy-tailed noise and weakly average smoothness.
method Normalized stochastic first-order methods with Polyak, multi-extrapolated, and recursive momentum.
result First-order oracle complexity results for finding approximate stochastic stationary points under heavy-tailed noise.
We demonstrate that, in the classical non-stochastic regret minimization problem with d d d decisions, gains and losses to be respectively maximized or minimized are fundamentally different. Indeed, by considering the additional sparsity assumption (at each stage, at most s s s decisions incur a nonzero outcome), we derive…
Optimal algorithms for Riemannian optimization with reduced complexity.
problem Stochastic optimization on Riemannian manifolds with limited data.
method Zeroth-order Riemannian Averaging Stochastic Approximation algorithms using Riemannian moving-average estimators and novel geometric conditions.
result Achieves optimal sample complexities for generating approximate first-order stationary solutions.
New algorithm achieves data-dependent regret bounds in MDPs with unknown transitions.
problem Achieving best-of-both-worlds guarantees with data-dependent regret bounds in MDPs with unknown transitions.
method Optimistic follow-the-regularized-leader algorithm with new optimistic Q-function estimators and transition bonus.
result First-order, second-order, and path-length bounds with polylog(T) regret in the stochastic regime.
Unified approach for first-order methods with Markovian noise in stochastic optimization and variational inequalities.
problem Stochastic optimization problems with Markovian noise.
method Unified theoretical analysis of first-order gradient methods using randomized batching and multilevel Monte Carlo.
result Optimal (linear) dependence on the mixing time of the noise sequence, eliminating previous limiting assumptions.
Certified algorithms optimize functions with varying costs, providing error bounds.
problem Optimizing functions with varying evaluation costs and error bounds.
method Formalized as a min-max game, proposed certified MFDOO algorithm with cost complexity bound.
result Proposed certified MFDOO algorithm has near-optimal cost complexity for Lipschitz functions.
Improved Local SGD convergence for general convex objectives with bounded second-order heterogeneity.
problem Understanding when and why Local SGD outperforms alternatives in distributed optimization.
method Established improved convergence guarantees for Local SGD on general convex objectives under bounded second-order heterogeneity.
result Upper bounds for Local SGD are nearly tight, providing a sharper convergence theory.
Paper develops a TR-SSQP method for noisy optimization with heavy-tailed noise.
problem Optimization problems with stochastic objectives and heavy-tailed noise.
method Trust-Region Stochastic Sequential Quadratic Programming (TR-SSQP) method.
result Achieves high-probability first-order and second-order stationarity bounds for heavy-tailed noise.
Paper proves higher-order flow matching preserves optimality in generative modeling.
problem Theoretical guarantees for higher-order flow matching in generative modeling.
method Neural network approximations with controlled depth, width, and sparsity.
result Proves worst case optimality for second-order flow matching.
First order methods can take extremely long to find global minima of non-convex functions.
problem Finding global minimizers of non-convex functions.
method Designing a family of non-convex functions and using statistical lower bounds for parameter estimation.
result First order methods can take exponential time to converge to a global minimizer.
Study optimizes zero-order strongly convex function minimization with higher order smoothness.
problem Optimizing a strongly convex function with noisy evaluations.
method Randomized approximation of projected gradient descent with smoothing kernel.
result Upper bounds and minimax lower bounds for the algorithm, showing near-optimality.
New algorithm reduces regret in infinite MDPs with optimal variance-dependent bounds.
problem Infinite horizon MDPs lack optimal algorithms with low regret.
method Developed a UCB-style algorithm for average-reward and γ-regret.
result Achieved optimal variance-dependent regret bounds for both objectives.
Many practitioners who use the EM algorithm complain that it is sometimes slow. When does this happen, and what can be done about it? In this paper, we study the general class of bound optimization algorithms - including Expectation-Maximization, Iterative Scaling and CCCP - and their relationship to direct optimizatio…
Method solves complex optimization problems with high probability bounds.
problem Nonlinear equality constrained stochastic optimization problems.
method Step-search sequential quadratic programming method.
result High-probability bound on iteration complexity for first-order stationarity.
Optimal simple regret bound for Gaussian Process bandits.
problem Sequential optimization of expensive-to-evaluate functions.
method Proved a bound on simple regret for pure exploration algorithms.
result Order optimal bound on simple regret for Gaussian Process bandits.
New inequalities help optimize first-order algorithms for statistical risk analysis.
problem Optimizing first-order iterative algorithms for statistical risk analysis.
method Introducing basic inequalities that connect implicit and explicit regularization.
result The basic inequalities translate the number of iterations into an effective regularization coefficient.
New methods bound estimation error in high-dimensional statistical problems.
problem Fundamental limits of first order methods in high-dimensional estimation.
method Introduces general first order methods for high-dimensional regression and low-rank matrix estimation.
result Derives optimal lower bounds on estimation error for these methods.
SOAR improves deep networks' robustness against adversarial examples.
problem Improving deep neural networks' robustness against adversarial examples.
method Formulated adversarial robustness problem under robust optimization framework, approximated loss function using second-order Taylor series expansion.
result SOAR significantly improves robustness of networks against adversarial perturbations.
New RL method handles large state-action spaces with complex models.
problem Complex models and large state-action spaces in reinforcement learning.
method π-KRVI, an optimistic modification of least-squares value iteration using kernel ridge regression.
result First order-optimal regret guarantees under general settings, improving over state of the art.
Unified framework for learning flexible probabilistic programs using DPP and PAC-Bayes bounds.
problem Learning and generalizing from complex probabilistic models.
method Unified DPP representation and PAC-Bayes bounds for stochastic programs.
result Improved performance and generalization prediction using flexible DPP model representations and learned complexity measures.
New bounds on optimal transport regularization show faster convergence rates than previously known.
problem Understanding the localization rate of Quadratically Regularized Optimal Transport (QOT) optimizers.
method Established lower bounds and derived mean-squared deviation controls for QOT optimizers.
result Lower bound of support concentration rate ε 1 d + 2 \varepsilon^{\frac{1}{d+2}} ε d + 2 1 in directed Hausdorff distance. New algorithm reduces regret in bandit optimization for high-dimensional data.
problem Optimizing decisions in uncertain environments with high-dimensional data.
method Inspired by online Newton step, proposes a simple and efficient BCO algorithm.
result Achieves optimal regret bounds for κ κ κ -convex functions. GN algorithm solves batched bandit for nondegenerate functions near-optimally.
problem Batched bandit learning for nondegenerate functions.
method Introduces Geometric Narrowing (GN) algorithm with a O ~ ( A + d T ) \widetilde{\mathcal{O}} ( A_{+}^d \sqrt{T} ) O ( A + d T ) regret bound and O ( log log T ) \mathcal{O} (\log \log T) O ( log log T ) batches. result GN achieves near optimal regret with minimal number of batches.
Unified bounds for iterative algorithms with Gaussian data matrices.
problem Establishing non-asymptotic bounds for iterative algorithms with Gaussian data.
method Explicit coupling between iterates and Gaussian process with deterministic covariance.
result Tight, dimension-free bounds for generalized first-order methods.
The thesis clarifies when local updates outperform centralized methods in heterogeneous data environments.
problem Understanding when local updates are more effective than centralized or mini-batch methods in distributed optimization.
method Fine-grained consensus-error-based analysis framework, focusing on bounded second-order heterogeneity and third-order smoothness.
result Local updates outperform centralized or mini-batch methods under realistic models of data heterogeneity.
Paper analyzes LSA algorithm bias and error bounds with RR extrapolation.
problem Analyzing bias and high-order error bounds of LSA with Markovian noise.
method Polyak-Ruppert averaging, linearization, Richardson-Romberg extrapolation.
result RR extrapolation effectively cancels the leading bias term.
Data processing inequalities link Fisher information to local differential privacy constraints.
problem Understanding how Fisher information scales with local differential privacy constraints.
method Developed data processing inequalities for Fisher information under local differential privacy.
result Implications for private estimation with optimal bounds and error rates.
We introduce a new recursive aggregation procedure called Bernstein Online Aggregation (BOA). The exponential weights include an accuracy term and a second order term that is a proxy of the quadratic variation as in Hazan and Kale (2010). This second term stabilizes the procedure that is optimal in different senses. We…
Optimization rates improved for manifolds with bounded geometry.
problem Optimizing functions on manifolds with bounded geometry.
method Riemannian gradient descent and dynamic trivialization algorithm.
result Curvature-dependent convergence rates computed explicitly for common manifolds.
We study the information-theoretic lower bound of the sample complexity of the correct recovery of diffusion network structures. We introduce a discrete-time diffusion model based on the Independent Cascade model for which we obtain a lower bound of order Ω ( k log p ) Ω(k \log p) Ω ( k log p ) , for directed graphs of p p p nodes, and at most k k k …
This paper achieves first-order regret bounds in reinforcement learning with large state spaces.
problem Achieving first-order regret bounds in reinforcement learning with large state spaces.
method Developed a novel robust self-normalized concentration bound based on the robust Catoni mean estimator.
result Obtained regret bounds scaling as O ~ ( d 3 H 3 ⋅ V 1 ⋆ ⋅ K + d 3.5 H 3 log K ) \widetilde{\mathcal{O}}(\sqrt{d^3 H^3 \cdot V_1^\star \cdot K} + d^{3.5}H^3\log K ) O ( d 3 H 3 ⋅ V 1 ⋆ ⋅ K + d 3.5 H 3 log K ) . The paper analyzes the efficiency of gradient estimation methods in noisy function evaluations.
problem Estimating gradients of smooth functions using noisy function evaluations.
method Information-theoretic lower bounds and finite difference method analysis.
result The finite difference method is not minimax optimal, suggesting room for improvement in gradient estimation.
This paper analyzes convergence of RMSProp and Adam in non-convex optimization with tight complexity bounds.
problem Analyzing convergence of RMSProp and Adam in non-convex optimization with relaxed assumptions.
method Developed new convergence analyses for RMSProp and Adam, considering adaptive learning rates and affine noise variance.
result RMSProp and Adam converge to ε-stationary points with iteration complexities of O(ε^(-4)) under proper hyperparameters.
Optimal algorithms for online convex optimization with random order.
problem Online convex optimization with random order and non-convex loss functions.
method Stochastic gradient descent and algorithmic stability analysis.
result Achieves optimal bounds and significantly outperforms previous methods.
In linear stochastic bandits, it is commonly assumed that payoffs are with sub-Gaussian noises. In this paper, under a weaker assumption on noises, we study the problem of \underline{lin}ear stochastic {\underline b}andits with h{\underline e}avy-{\underline t}ailed payoffs (LinBET), where the distributions have finite…
New bounds show current methods overestimate system parameter errors.
problem Current bounds overestimate parameter errors in system identification.
method Utilized asymptotic normality and second-order decomposition.
result Obtained finite-sample bounds matching optimal rates up to constants.
Optimal transport theory characterizes convex order between probability measures.
problem Characterizing convex order between probability measures using optimal transport.
method Quantitative bounds on optimal transport, infimum of functionals over 1-Lipschitz functions.
result Two measures are in convex order if and only if a specific cost functional inequality holds.
Quadratic memory is essential for optimal convex optimization queries.
problem Optimal query complexity for convex optimization and feasibility problems.
method Lower bounds on query complexity for convex optimization and feasibility problems.
result Center-of-mass algorithms are Pareto-optimal for both convex optimization and feasibility problems.