New online conformal prediction methods minimize strongly adaptive regret and achieve near-optimal coverage.
problem Uncertainty quantification in online settings with changing data distributions.
method Developed new online conformal prediction methods that minimize strongly adaptive regret.
result Achieve near-optimal strongly adaptive regret and approximately valid coverage.
New algorithm improves online learning with expert advice and metric learning.
problem Improving online learning performance in changing environments.
method Parameter-free online learning algorithm using coin betting.
result Strongly adaptive regret bound improvement of at least sqrt(log(T)).
New meta algorithm improves adaptability in changing environments.
problem Adapting to changing environments in online learning.
method Derives a new parameter-free algorithm for the LEA problem, inspired by coin betting.
result Strongly-adaptive regret bound is log ( T ) \sqrt{\log(T)} log ( T ) better than other algorithms. SA algorithms control dynamic regret in non-stationary settings with strong convexity or exp-concavity.
problem Non-stationary Online Convex Optimization with dynamic regret control.
method Strongly Adaptive (SA) algorithms view dynamic regret as path variation of the comparator sequence.
result SA algorithms achieve i l d e O ( T V T ∨ log T ) ilde O(\sqrt{TV_T} \vee \log T) i l d e O ( T V T ∨ log T ) and i l d e O ( d T V T ∨ d log T ) ilde O(\sqrt{dTV_T} \vee d\log T) i l d e O ( d T V T ∨ d log T ) dynamic regret for strongly convex and exp-concave losses, respectively. New algorithms minimize dynamic regret for strongly convex losses.
problem Minimizing dynamic regret for strongly convex losses.
method Developed Strongly Adaptive algorithms exploiting KKT conditions.
result Achieved near optimal dynamic regret of O ( d 1 / 3 n 1 / 3 e x t T V [ u 1 : n ] 2 / 3 ∨ d ) O(d^{1/3} n^{1/3} ext{TV}[u_{1:n}]^{2/3} \vee d) O ( d 1/3 n 1/3 e x t T V [ u 1 : n ] 2/3 ∨ d ) . Paper analyzes and improves adaptive gradient methods for optimization.
problem Improving optimization methods for deep neural networks.
method Analyzes and proposes variants of RMSProp and Adagrad for online convex optimization.
result Proposes SC-Adagrad and SC-RMSProp with logarithmic regret bounds for strongly convex functions.
Unintended effects from scaling neural network outputs with adaptive learning rates.
problem Adaptive learning rate optimization's behavior is altered by output scaling, leading to misinterpretation.
method Presented a modified optimization algorithm to mitigate unintended effects.
result Adaptive learning rate's effectiveness is significantly impacted by output scaling, especially for small scaling factors.
Improved SHB method for faster convergence on strongly-convex quadratics.
problem Understanding and improving the theoretical and practical advantages of SHB.
method Noise-adaptive multi-stage algorithm for SHB with accelerated convergence.
result SHB can achieve accelerated convergence with larger mini-batch sizes.
New proof confirms periodic orbit conjecture for Eulerisable flows.
problem Periodic orbit conjecture for non-vanishing vector fields on closed manifolds.
method Characterization of Eulerisable flows and use of strongly adapted one-forms.
result Periodic orbit conjecture holds for Eulerisable flows.
Adaptive SGD learns optimal batch size for strong convex functions.
problem Finding optimal batch size for SGD in practice.
method Adaptive SGD method that learns optimal batch size.
result Adaptive SGD exhibits nearly optimal performance in experiments.
Adaptive step sizes improve optimization for convex and nonconvex problems.
problem Optimizing functions that are not strongly convex.
method Bridge nonconvex and strongly convex problems via regularization, then apply Barzilai-Borwein step sizes with SARAH.
result Regularized SARAH methods achieve better complexity in nonconvex problems.
Investigates connections adapted to a holomorphic Lie group action on bundles.
problem Finding connections adapted to a Lie group action on bundles.
method Analyzes connections on principal H H H -bundles over complex manifolds with holomorphic actions of Lie groups. result Identifies conditions for connections to be adapted to a given G G G -connection. Upper and lower bounds derived for online learning with graph-structured feedback against adaptive adversaries.
problem Online learning with graph-structured feedback against adaptive adversaries.
method Analysis of Exp3 algorithm variants and lower bounds for specific adversary models.
result Upper bounds of O ~ ( T 2 / 3 ) \widetilde O(T^{2/3}) O ( T 2/3 ) and O ~ ( T 3 / 4 ) \widetilde O(T^{3/4}) O ( T 3/4 ) for strongly-observable and weakly-observable graphs, respectively. Modified dynamical systems retain Turing universality.
problem Embedding Turing machines into dynamical systems.
method Exploring flows with adapted 1-forms and homogeneity.
result Even slight modifications can lead to Turing universality.
This paper analyzes adaptive gradient algorithms for better performance in ill-conditioned problems.
problem Poor performance of standard stochastic gradient algorithms in ill-conditioned problems.
method Non-asymptotic analysis of adaptive gradient algorithms (Adagrad and Stochastic Newton) for strongly convex objectives.
result Theoretical analysis and adaptation to practical applications like linear regression and regularized GLM.
In this work we introduce a new optimisation method called SAGA in the spirit of SAG, SDCA, MISO and SVRG, a set of recently proposed incremental gradient algorithms with fast linear convergence rates. SAGA improves on the theory behind SAG and SVRG, with better theoretical convergence rates, and has support for compos…
ANIL adapts only a subset of parameters, reducing computational cost.
problem Efficiently adapt model parameters in meta-learning.
method Adapts only a small subset of parameters in the inner loop of ANIL.
result Theoretical convergence and computational complexity analysis for ANIL.
Characterizes Anosov flows via contact geometry.
problem Understanding Anosov 3-flows through contact geometry.
method Investigates interactions with Reeb dynamics and proves a technical theorem.
result Space of adapted geometries homotopy equivalent to Anosov flows.
Optimal online linear regression in dynamic environments using discounted Vovk-Azoury-Warmuth forecaster.
problem Achieving optimal performance in dynamic online linear regression without prior knowledge.
method Developed a discounted variant of the Vovk-Azoury-Warmuth forecaster to achieve optimal dynamic regret guarantees.
result Achieved dynamic regret of the form $O\left(d\log(T)\vee \sqrt{dP_{T}^γ(\vec{u})T}
ight)$ , with a learnable discount factor.
New algorithm reduces TV-denoising to adaptive online learning.
problem Estimating TV-bounded functions from noisy samples.
method Deep connection to Strongly Adaptive online learning; O ( n log n ) O(n \log n) O ( n log n ) time algorithm. result Near minimax optimal rate of O ( n 1 / 3 C n 2 / 3 ) O(n^{1/3}C_n^{2/3}) O ( n 1/3 C n 2/3 ) under squared error loss. New adaptive models improve prediction accuracy with missing data.
problem Improving prediction accuracy with missing data entries.
method Adaptive optimization approach, learning imputation and regression simultaneously.
result 2-10% improvement in out-of-sample accuracy in strongly non-random missing data settings.
The study shows strong formality in certain complex manifolds.
problem Investigating strong formality in complex manifolds.
method Adapting s s s -strong formality from Fernandez and Muñoz to the pluripotential setting. result Compact Kähler manifolds and generalized complete intersections are strongly formal.
Random extrapolation speeds up coordinate descent for sparse and dense data.
problem Efficiently solving primal-dual coordinate descent for sparse and dense data.
method Adapts to sparsity and uses large step sizes for dense data, proving linear convergence under metric subregularity.
result Linear convergence under metric subregularity and optimal sublinear convergence rates in general convex-concave problems.
New algorithm tackles heterogeneous curvature in online convex optimization.
problem Adversarial bandit convex optimization with varying curvature.
method Developed an adaptive algorithm that learns curvature on the fly.
result Achieves optimal regret bounds even with heterogeneous curvature.
New methods optimize functions faster with less gradient accuracy needed.
problem Optimizing complex functions with limited gradient accuracy.
method Hessian averaging and adaptive gradient sampling methods.
result Improved convergence rates for various function types.
Adaptive personalized federated learning improves local model personalization.
problem Maximizing global model performance limits local model personalization.
method APFL algorithm trains local models while contributing to global model, with optimal mixing parameter and communication-efficient optimization.
result Demonstrates effectiveness of personalization schema and correctness of generalization theories.
TiAda adapts adaptive gradient methods for nonconvex minimax optimization.
problem Nonconvex minimax optimization challenges in achieving convergence.
method TiAda is a time-scale adaptive GDA algorithm for nonconvex minimax optimization.
result TiAda achieves near-optimal complexities in deterministic and stochastic settings.
New theory for clustering in geometric and adaptive settings.
problem Clustering in non-Euclidean spaces and adaptive parameters.
method Asymptotic theory for k k k -means and related methods. result Strong consistency and asymptotic limit theorems for various clustering procedures.
Three adaptive methods improve financial forecasting and portfolio management.
problem Improving financial forecasting and portfolio management in volatile markets.
method Dynamic Model Selection (DMS), Adaptive Ensemble (AE), Dynamic Asset Allocation (DAA).
result Adaptive methods outperform long-only benchmarks in US market returns.
New method achieves both universality and adaptivity in online convex optimization.
problem Achieve optimal regret guarantees without prior knowledge of function curvature.
method Introduces UniGrad, a novel approach that achieves both universality and adaptivity.
result Achieves universal regret guarantees that adapt to gradient variation.
Introduces CSLC models to bridge deep generative models and classical algorithms.
problem Mode collapse and memorization issues in deep generative models and restrictive assumptions in classical algorithms.
method Introduces conditionally strongly log-concave (CSLC) models, factorizing data distribution into strongly log-concave conditional distributions.
result Efficient parameter estimation and sampling algorithms with theoretical guarantees for non-log-concave data distributions.
Adaptive step-size improves optimization in complex geometries.
problem Optimizing functions with non-Euclidean geometries.
method Adaptive step-size strategy for optimization algorithms.
result Guaranteed convergence for Adaptive Conditional Gradient Descent.
A Kernel Adaptive Metropolis-Hastings algorithm is introduced, for the purpose of sampling from a target distribution with strongly nonlinear support. The algorithm embeds the trajectory of the Markov chain into a reproducing kernel Hilbert space (RKHS), such that the feature space covariance of the samples informs the…
Compactifies SU(2) monopole moduli spaces, proving part of Sen's conjecture.
problem Proving part of Sen's conjecture for L2 cohomology of moduli spaces.
method Compactifications of moduli spaces of SU(2) monopoles on R3, leading order asymptotic expansions of hyperKaehler metrics.
result Proves part of Sen's conjecture for the L2 cohomology of strongly centered moduli spaces.
Adaptive sampling method reduces variance in stochastic optimization.
problem Reducing variance in stochastic optimization with limited gradient computations.
method Adaptive increase in sample size based on inner product test.
result Algorithm converges globally on nonconvex functions and linearly on strongly convex functions.
Improves MCMC performance with adaptive affine transformations.
problem Improving the performance of Markov Chain Monte Carlo samplers.
method Adaptive learning of bijective affine transformations during sampling.
result Adaptive affine transformations improve the quality of samples at low computational cost.
We consider the problems of detection and localization of a contiguous block of weak activation in a large matrix, from a small number of noisy, possibly adaptive, compressive (linear) measurements. This is closely related to the problem of compressed sensing, where the task is to estimate a sparse vector using a small…
In this paper, we introduce Adaptive Cluster Lasso(ACL) method for variable selection in high dimensional sparse regression models with strongly correlated variables. To handle correlated variables, the concept of clustering or grouping variables and then pursuing model fitting is widely accepted. When the dimension is…
Universal algorithm minimizes adaptive regret for various convex functions.
problem Minimizing adaptive regret in changing environments for multiple convex functions.
method Borrowing MetaGrad's idea of multiple learning rates and using sleeping experts.
result First universal algorithm for minimizing adaptive regret of convex functions.
New algorithm reduces adaptive regret without projections.
problem Computational expense of projections in online convex optimization.
method Lazy gradient-based algorithm with set-membership computations.
result Near-optimal adaptive regret bounds for general convex functions.
FDN improves probabilistic regressors' adaptability to distribution shifts.
problem Overconfidence in modern probabilistic regressors under distribution shift.
method FDN uses input-conditioned distributions over network weights, trained with a Monte Carlo beta-ELBO objective.
result FDN produces predictive mixtures whose dispersion adapts to the input, providing shift-aware uncertainty.
Gaussian Mixture Models (GMM) have found many applications in density estimation and data clustering. However, the model does not adapt well to curved and strongly nonlinear data. Recently there appeared an improvement called AcaGMM (Active curve axis Gaussian Mixture Model), which fits Gaussians along curves using an …
Researchers adaptively analyze market regimes to reveal investor behavior shifts.
problem Market relationships shift across different regimes, affecting investor behavior.
method Combining Kalman filtering, Markov-switching, and asymmetric response estimation.
result Foreign investors' predictive power increases during crises, while individual investors react more strongly to positive shocks.
Adaptive method improves numerical solution of Cox-Ingersoll-Ross model.
problem Approximating solutions to the Cox-Ingersoll-Ross model efficiently.
method Path-bounded timestepping with hybrid approach, including a backstop method.
result The adaptive method is strongly convergent, with strong error control.
An adaptive algorithm optimizes resource allocation with diminishing returns.
problem Sequential resource allocation with diminishing returns.
method Adaptive stochastic optimization algorithm that minimizes regret.
result Optimizes cumulative reward with optimal rates for strongly-concave functions and classical multi-armed bandit rates.
ESS improves MCMC efficiency for correlated & multimodal distributions.
problem Slice Sampling's sensitivity to initial length scale and difficulty with correlated distributions.
method Adaptive tuning and parallel walkers for efficient sampling.
result ESS improves efficiency by more than an order of magnitude on correlated distributions.
New insights show NAG and FISTA converge linearly without knowing strong convexity modulus.
problem Understanding linear convergence of NAG and FISTA without strong convexity modulus knowledge.
method High-resolution ODE framework, dynamically adapting kinetic energy coefficient.
result NAG and FISTA demonstrate linear convergence without requiring strong convexity modulus knowledge.
A new gradient method LAG reduces communication in distributed learning.
problem Reducing communication in distributed machine learning.
method Skip gradient calculations for slowly-varying gradients using simple rules.
result Communication reduction with theoretical and empirical support.