Polyak step size GD reaches final radius of convergence after log iterations.
problem Statistical and computational complexities of Polyak step size GD.
method Generalized smoothness and Lojasiewicz conditions, stability of gradients.
result Polyak step size GD reaches final statistical radius of convergence after logarithmic number of iterations.
The paper interprets learned step sizes in deep-unfolded gradient descent.
problem Intuitive interpretation of learned non-constant step sizes in deep-unfolded gradient descent.
method Theoretical analysis and optimization of spectral radius.
result Chebyshev steps achieve the lower bound of convergence rate for first-order methods.
Paper develops an online learning algorithm for functional data models.
problem Recovering slope functions or predictors in functional data models.
method Online regularized learning algorithm in reproducing kernel Hilbert spaces with polynomially decaying step-size.
result Established fast convergence rates for estimation error without capacity assumption.
Develops a generalized version of Chung's Lemma for stochastic optimization methods.
problem Establishing asymptotic convergence rates for stochastic optimization methods under various step size rules.
method Generalized version of Chung's Lemma for a broader family of step size rules.
result Demonstrates tight non-asymptotic convergence rates for various stochastic methods.
Paper analyzes online learning without regularization, proving strong convergence.
problem Online learning without explicit regularization terms.
method Stochastic gradient descent in RKHS with polynomially decaying step sizes.
result Strong convergence of the last iterate in RKHS norm with polynomial step sizes.
Proposes an exponentially increasing step-size for faster parameter estimation in statistical models.
problem Slow convergence of gradient descent in locally convex loss functions.
method Exponentially increasing step-size in gradient descent algorithm.
result Converges linearly to optimal solution under homogeneous assumptions.
The paper analyzes and validates two step size schedules for SGD: exponential and cosine, proving their adaptivity and performance.
problem The variability of SGD performance due to step size choice.
method Analysis and empirical evaluation of exponential and cosine step sizes.
result Exponential and cosine step sizes are adaptive to noise and achieve optimal performance without tuning hyperparameters.
Gradient descent on logistic loss converges to the maximum-margin separator for separable data.
problem Understanding the convergence of gradient descent on separable datasets with specific loss functions.
method Analysis of gradient descent on linear models with super-polynomially tailed losses.
result For separable datasets, gradient descent converges to the maximum-margin separator for losses with super-polynomial tails, but not for heavier tails.
Temporal Difference Learning analysis under non-i.i.d. data and nonlinear approximation.
problem Finite-sample behavior of TD(0) under non-i.i.d. data and nonlinear approximation.
method High-probability, finite-sample analysis of vanilla TD(0) on polynomially mixing Markov data, assuming Holder continuity and bounded generalized gradients.
result Bounds on the convergence rate of TD(0) with high probability, matching known i.i.d. rates and holding even with nonstationary initialization.
Chebyshev steps improve convergence in deep-unfolded gradient descent.
problem Improving convergence speed in iterative algorithms.
method Introducing Chebyshev steps to bound convergence rate of gradient descent.
result Chebyshev steps lead to asymptotically optimal convergence rate.
Study shows how mini-batch GD with random reshuffling affects least squares regression dynamics.
problem Analyzing the error dynamics of mini-batch GD with random reshuffling for least squares regression.
method Represented training and generalization errors through a sample cross-covariance matrix Z, compared with sample covariance matrix of original features X, and used linear scaling rule for analysis.
result Mini-batch GD with random reshuffling exhibits subtle step-size dependence not detectable by gradient flow analysis, converging to a limit dependent on the step size.
Geometric step decay schedules improve stochastic algorithms' convergence on sharp nonconvex problems.
problem Convergence of stochastic algorithms on sharp nonconvex problems.
method Geometric step decay schedule applied to stochastic algorithms.
result Geometric step decay schedules lead to local linear convergence rates for sharp nonconvex problems.
Gradient descent with growing learning rate enables learning non-linear features in neural networks.
problem Learning non-linear features in two-layer neural networks.
method Using gradient descent with a learning rate that grows with the sample size.
result Multiple rank-one components emerge, each corresponding to a specific polynomial feature.
Reward-poisoning attacks can force RL agents to learn bad policies, and we categorize and quantify their feasibility.
problem Reward-poisoning attacks can manipulate RL agents to learn undesirable policies.
method Categorize attacks by infinity-norm constraint, provide thresholds for feasibility, and develop adaptive attack strategies.
result Adaptive reward-poisoning attacks can achieve the nefarious policy in polynomial steps, while non-adaptive attacks require exponential steps.
Guarantees for tuning step size using meta-gradient descent.
problem Choosing optimal parameters for optimization algorithms.
method Learning-to-learn approach with meta-gradient descent.
result Meta-gradient can explode/vanish, requiring careful meta-objective design.
Improved learning rate schedule for least squares regression.
problem Achieving optimal convergence rates for least squares regression.
method Step Decay schedule with geometrically decaying learning rates.
result Final iterate behavior with Step Decay schedules is off the minimax rate by only log factors.
Polynomial-time reachability for LTI systems with TLL NN controllers is achieved.
problem Bounding the reachable set of LTI systems controlled by TLL NN controllers.
method Polynomial-time computation of exact one-step reachable set and tight bounding box via two methods.
result Exact reachability computation in polynomial time for TLL NN controllers.
Warm starts improve variational quantum algorithms by avoiding barren plateaus.
problem Barren plateaus in variational quantum algorithms limit scaling.
method Exploring warm starts in iterative variational methods for quantum circuits.
result Warm starts can lead to substantial gradients in small regions, suggesting trainability.
Study of two-layer NNs under Gaussian mixtures data, proving polynomial models equivalent to neural networks.
problem Training and generalization performance of two-layer NNs under structured Gaussian mixture data.
method Asymptotic analysis of two-layer NNs after one gradient descent step under Gaussian mixture data assumption.
result High-order polynomial models equivalent to nonlinear neural networks under certain conditions.
Study convergence of simulated annealing in continuous and discrete settings.
problem Analyzing convergence rate of simulated annealing methods.
method Apply Eyring-Kramers law to prove polynomial decay of tail probabilities.
result Explicit rate of convergence for continuous and discrete simulated annealing.
Implicit Q-learning and SARSA adjust step-sizes automatically, improving stability and performance.
problem Numerical instability and slow progress in Q-learning and SARSA due to step-size calibration.
method Reformulate iterative updates as fixed-point equations, scaling step-sizes inversely with feature norms.
result Implicit methods maintain stability over broader step-size ranges and achieve comparable convergence rates.
Neural circuits integrate continuous dynamics efficiently.
problem Efficiently integrating continuous neural dynamics for simulation and learning.
method Compact neural circuits for Runge-Kutta and Adams-Bashforth-Moulton methods.
result Equivalence of neural and numerical integration for polynomial systems.
In this paper, we consider unregularized online learning algorithms in a Reproducing Kernel Hilbert Spaces (RKHS). Firstly, we derive explicit convergence rates of the unregularized online learning algorithms for classification associated with a general gamma-activating loss (see Definition 1 in the paper). Our results…
Improved computational complexity in statistical models using second-order information.
problem Polynomial convergence of gradient descent in singular statistical models.
method Normalized Gradient Descent (NormGD) algorithm with second-order information.
result NormGD reaches final statistical radius in logarithmic iterations of n. TIDBD adapts TD step-sizes for better performance.
problem Finding optimal step-sizes for TD learning.
method Generalizes IDBD to TD learning, adapting step-sizes per feature.
result TIDBD outperforms other TD methods in various tasks.
The paper analyzes fixed step-size SA schemes on Riemannian manifolds.
problem Developing efficient algorithms for optimization on curved spaces.
method Fixed step-size stochastic approximation schemes in a Riemannian framework.
result The schemes converge to the solution as the step-size approaches zero.
Independent component analysis (ICA) aims at decomposing an observed random vector into statistically independent variables. Deflation-based implementations, such as the popular one-unit FastICA algorithm and its variants, extract the independent components one after another. A novel method for deflationary ICA, referr…
Adaptive step size improves online learning performance.
problem Tedious tuning of step size in online learning algorithms.
method Adapt step size by gradient descent on the step size itself.
result Online adaptation reduces performance variability.
New method shows AdaGrad converges globally to neural network minima.
problem Convergence of adaptive gradient methods for neural networks.
method Proposed adaptive gradient method for over-parameterized neural networks.
result Converges to global minimum in polynomial time for two-layer networks.
The CSA-ES is an Evolution Strategy with Cumulative Step size Adaptation, where the step size is adapted measuring the length of a so-called cumulative path. The cumulative path is a combination of the previous steps realized by the algorithm, where the importance of each step decreases with time. This article studies …
New algorithm improves stability of optimization algorithms by adapting step-size.
problem Optimization algorithms' effectiveness is sensitive to step-size hyperparameters.
method Adapts NGN step-size method with momentum to enhance stability.
result Achieves convergence rate of O(1/√K) without restrictive assumptions.
New algorithm converges geometrically fast with different step-sizes.
problem Distributed optimization with varying step-sizes.
method Adapt-Then-Combine (ATC) variation of DIGing algorithm.
result Geometric convergence with uncoordinated step-sizes.
Stagewise training outperforms vanilla SGD in accelerating convergence and testing error reduction.
problem Improving the convergence rate of SGD for neural networks.
method Stagewise training strategy with a geometrically decreasing step size, compared to vanilla SGD with a polynomially decaying step size.
result Stagewise training achieves faster convergence and testing error reduction compared to vanilla SGD under the Polyak-Łojasiewicz condition.
AutoStep MCMC adapts step size locally for better sampling efficiency.
problem Challenging step size selection for complex, multiscale targets.
method AutoStep MCMC uses a locally adaptive step size for involutive proposals.
result AutoStep MCMC is π-invariant, irreducible, and aperiodic.
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.
This paper optimizes step sizes for faster convergence in sparse coding.
problem Improving convergence rate of ISTA for sparse coding.
method Learning step sizes for ISTA using neural networks.
result A simple step size strategy can improve convergence rate of ISTA.
New algorithm approximates maximum of certain distributions on subsets.
problem Finding maximum of distributions on subsets.
method Connection between sampling and optimization via exchange inequalities and local random walks.
result Simple nearly-optimal approximation algorithm for MAP inference.
Paper proposes a new method for choosing step sizes in SGD and SVRG.
problem Choosing an appropriate step size in stochastic gradient descent.
method Introduces Barzilai-Borwein method for automatically computing step sizes.
result SVRG-BB converges linearly for strongly convex objective functions.
We analyze constant step-size and iterate averaging in linear stochastic approximation algorithms.
problem Policy evaluation in reinforcement learning using temporal difference algorithms.
method Constant step-size and Polyak-Ruppert averaging of iterates.
result MSE decays as O(1/t) for a range of constant step-sizes under certain conditions.
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.
Negative step sizes improve second-order methods for neural networks.
problem Second-order methods discard negative curvature, limiting their effectiveness.
method Introduce negative step sizes in second-order methods combined with Wolfe line search.
result Negative step sizes lead to global convergence and improved performance.
New convergence results for NGVI with various step sizes and sample sizes.
problem Understanding convergence of stochastic NGVI for various schedules.
method Projected stochastic NGVI for exponential family variational distributions.
result Geometric convergence and $\mathcal{O}\left(\frac{1}{T^ρ}
ight)$ rates for different schedules.
New optimal step sizes and mini-batch sizes for SAGA.
problem Finding optimal step sizes and mini-batch sizes for SAGA.
method Provided closed-form expressions for expected smoothness constant and suggested new step sizes and mini-batch sizes.
result Total complexity of SAGA decreases linearly with mini-batch size up to an optimal value.
Study on RF regression with SGD shows double descent phenomenon.
problem Understanding generalization in RF models trained with SGD.
method Precise non-asymptotic error bounds derived for RF regression under constant and polynomial-decay step-size SGD.
result RF regression generalizes well for interpolation learning and exhibits double descent behavior.
Stochastic variational inference fails in complex Bayesian networks, requiring careful step size scaling.
problem Applying stochastic variational inference to general Bayesian networks leads to convergence issues.
method Investigates the scaling of initial step sizes in natural gradient steps for exponential family approximations, and how these steps are affected in a general Bayesian network context.
result Practical convergence requires careful consideration of initial step sizes, as natural gradient steps thrown out with the bath water.
Adaptive step-size method improves compressed SGD performance in machine learning.
problem Communication bottleneck in distributed and decentralized optimization.
method Developed an adaptive step-size method for compressed SGD.
result Order-optimal convergence rates for various objective functions.
Proposes a neural network for learning step-size policies for L-BFGS optimization.
problem Optimizing step sizes for L-BFGS in large-scale problems.
method Neural network architecture using local iterate information, trained via stochastic optimization.
result Outperforms existing step size selection methods in training classifiers.
New SVRG and SARAH schemes reduce tuning effort for variance reduction.
problem Optimal performance of SVRG and SARAH requires tuning of parameters.
method Introduces Barzilai-Borwein step sizes, averaging, and adaptive inner loop length.
result Improves SVRG, SARAH, and BB variants' convergence rates and performance.