
MOReL : ModelBased Offline Reinforcement Learning
In offline reinforcement learning (RL), the goal is to learn a successfu...
read it

The Step Decay Schedule: A Near Optimal, Geometrically Decaying Learning Rate Procedure
There is a stark disparity between the step size schedules used in pract...
read it

On the insufficiency of existing momentum schemes for Stochastic Optimization
Momentum based stochastic gradient methods such as heavy ball (HB) and N...
read it

Leverage Score Sampling for Faster Accelerated Regression and ERM
Given a matrix A∈R^n× d and a vector b ∈R^d, we show how to compute an ϵ...
read it

Efficient Estimation of Generalization Error and BiasVariance Components of Ensembles
For many applications, an ensemble of base classifiers is an effective s...
read it

A Markov Chain Theory Approach to Characterizing the Minimax Optimality of Stochastic Gradient Descent (for Least Squares)
This work provides a simplified proof of the statistical minimax optimal...
read it

Accelerating Stochastic Gradient Descent
There is widespread sentiment that it is not possible to effectively uti...
read it

Parallelizing Stochastic Approximation Through MiniBatching and TailAveraging
This work characterizes the benefits of averaging techniques widely used...
read it

Submodular Hamming Metrics
We show that there is a largely unexplored class of functions (positive ...
read it
Rahul Kidambi
is this you? claim profile