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 9 of 9 for “"Mirror Descent"”.

  1. On sparse mirror descent

    Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2022-11-11 without embargo terms

    uiuc Repository record for On sparse mirror descent (opens in a new tab)

  2. Learning distributions with Particle Mirror Descent

    … to estimate posterior distribution, Particle Mirror Descent is appealing for its simplicity and flexibility. In this thesis we explore the applications of Particle Mirror Descent in both supervised and unsupervised learning. In the general classification problem with a parametric …

    uiuc Repository record for Learning distributions with Particle Mirror Descent (opens in a new tab)

  3. A Unified Approach to Controlling Implicit Regularization Using Mirror Descent

    … by the optimization algorithms, such as gradient descent (GD), something referred to as implicit regularization. In particular, it has been argued that GD tends to induce an implicit $\ell_2$-norm regularization in regression and classification problems. Despite significant progress in this space, …

    mit Repository record for A Unified Approach to Controlling Implicit Regularization Using Mirror Descent (opens in a new tab)

  4. Large-Scale Optimization Methods: Theory and Applications

    … thesis, we show the efficiency of coordinate descent (CD) and mirror descent (MD) methods in solving large-scale optimization problems. First, we investigate the convergence rate of the CD method with different coordinate selection rules. We present certain problem classes, for which …

    mit Repository record for Large-Scale Optimization Methods: Theory and Applications (opens in a new tab)

  5. On Solving Larger Games: Designing New Algorithms Adaptable to Deep Reinforcement Learning

    … Chapter 3 introduces Regularized Optimistic Mirror Descent (Reg-OMD), which provably converges to the Nash equilibrium (NE) linearly in last-iterate. Chapter 4 shows that algorithms based on regret decomposition enjoy best-iterate convergence to the NE. Chapter 5 proposes Q-value based Regret …

    mit Repository record for On Solving Larger Games: Designing New Algorithms Adaptable to Deep Reinforcement Learning (opens in a new tab)

  6. Explicit Regularization for Overparameterized Models

    … with explicit regularization, called Regularizer Mirror Descent (RMD). In the overparameterized regime, where the number of model parameters exceeds the size of data, RMD provably converges to a point “close” to a minimizer of the regularized objective. Additionally, RMD is computationally …

    mit Repository record for Explicit Regularization for Overparameterized Models (opens in a new tab)

  7. Geometric methods in computational optimal transport and high-dimensional inference

    … well as to Wasserstein barycenters. Third, the Mirror Sinkhorn algorithm is introduced, unifying mirror descent with matrix scaling in a single loop procedure for optimising convex functions over transport polytopes. For B-Lipschitz objectives, it is shown that the algorithm achieves an O B√δT …

    cambridge Repository record for Geometric methods in computational optimal transport and high-dimensional inference (opens in a new tab)

  8. Efficient learning of temporal dynamics with first-order methods

    … we develop a general framework, named “pseudo mirror descent”, to address the challenge in efficient handling of the positivity constraint when learning the intensity function of point processes. This framework greatly alleviates the burden of expensive projections required by existing …

    uiuc Repository record for Efficient learning of temporal dynamics with first-order methods (opens in a new tab)

  9. Designing Provably Convergent Algorithms from the Geometry of Data

    … optimisation. Utilising machine learning in the mirror descent algorithm, we significantly accelerate convex optimisation in the presence of similar data, defined by optimisation objective functions. (ii) We then consider when the regularisation term is instead implicitly defined by its proximal …

    cambridge Repository record for Designing Provably Convergent Algorithms from the Geometry of Data (opens in a new tab)