Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 8 of 8 for “"adaptive stepsize"”.
-
High-dimensional Limit of SGD for Adaptive Stepsize Algorithms and Diagonal Linear Networks
… rely on asymptotic regimes with vanishing stepsizes, leading to continuous-time approximations such as deterministic gradient flow or small-noise diffusion models, which often fail to capture the rich behavior observed in practice. This thesis develops a unified framework for describing …
-
The analysis and implementation of exponential almost Runge-Kutta methods for semilinear problems
… properties. We then design a numerical, adaptive stepsize, differential equation solver. The core of the solver is powered by EIs, in particular our new family of EAGLMs. Experiments are performed against a comprehensive suite of test problems, with emphasis on stiff systems of …
-
Reliable Approximate Solution of Systems of Delay Volterra Integro-differential Equations
… Runge-Kutta (CRK) method. We introduce an adaptive stepsize selection strategy resulting in an approximate solution whose associated defect (residual) satisfies certain properties that allow us to monitor the global error reliably and efficiently. We prove the classic and optimal …
-
Function Space Complementarity Methods for Optimal Control Problems
… due to the inevitable discretization errors. The adaptive stepsize selection is based on a problem-adapted norm that takes the special affine invariance of optimization problems into account. The proposed method is robust and able to solve intricate problems to high accuracy. The main advantage …
-
Application of numerical methods to dislocation modeling of creep
… climb mobilities. Dual-order, error-estimating, adaptive stepsize Runge-Kutta methods of Verner and of Fehlberg are used to solve the systems of ODE's. Predictions for the behavior of the stress exponent as a function of stress and model geometric parameters differ from predictions obtained using …
-
Combatting ill-conditioning and heavy-tailed noise in non-convex optimization
… stochastic gradient method with block-coordinate stepsizes that effectively combat imbalanced gradients and inhomogeneous noise. Finally, we propose a statistically efficient data-driven method that adapts to the heavy-tail distribution of the noise. In Chapter 2, we analyze a preconditioned …
-
Accelerated Inexact First-Order Methods for Solving Nonconvex Composite Optimization Problems
… inexact proximal point framework under different stepsizes and levels of inexactness. Following our developments for the unconstrained setting, we move to study two instances of a function-constrained NCO problem. The first instance comprises a set of linear set constraints, and we develop a …
-
Distributed Learning with Automated Stepsizes
Stepsizes for optimization problems play a crucial role in algorithm convergence, where the stepsize must undergo tedious manual tuning to obtain near-optimal convergence. Recently, an adaptive method for automating stepsizes was proposed for centralized optimization. However, this method is not …