Rohith Rongali

Generalization Guarantees on Data-Driven Tuning of Gradient Descent with Langevin Updates

This work proves the existence of an optimal hyperparameter configuration for which the LGD algorithm achieves the Bayes'optimal solution for squared loss on regression tasks, and shows empirical evidence of the success of LGD and the meta learning procedure for few-shot learning on linear and logistic regression using...