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 5 of 5 for “"Overparameterized models"”.

  1. Explicit Regularization for Overparameterized Models

    … 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 efficient and imposes virtually no …

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

  2. One-Pass Learning via Bridging Orthogonal Gradient Descent and Recursive Least-Squares

    … datapoints. Motivated by the increasing use of overparameterized models, we develop Orthogonal Recursive Fitting (ORFit), an algorithm for one-pass learning which seeks to perfectly fit every new datapoint while changing the parameters in a direction that causes the least change to the …

    mit Repository record for One-Pass Learning via Bridging Orthogonal Gradient Descent and Recursive Least-Squares (opens in a new tab)

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

    … understanding the generalization performance of overparameterized models and the effect of optimization algorithms on it has become an increasingly popular question. In particular, there has been substantial effort to characterize the solutions preferred by the optimization algorithms, such as …

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

  4. Methods for Generalization Under Distribution Shift

    … distribution shifts is collecting and training models on large datasets. This work offers two more principled frameworks that enable machine learning models to generalize effectively to out-of-distribution scenarios without sacrificing the power of modern overparameterized models. The first …

    mit Repository record for Methods for Generalization Under Distribution Shift (opens in a new tab)

  5. Learning to classify images without explicit human annotations

    … labels from workers, and third training a large, overparameterized deep neural network on the clean examples. The second, manual labeling step is often the most expensive one as it requires manually going through all examples. In this thesis we propose to i) skip the manual labeling step entirely, …

    rice Repository record for Learning to classify images without explicit human annotations (opens in a new tab)