Sparsifying Parametric Models with L0 Regularization

TL;DR

This paper introduces an educational approach to sparsify parametric models using L0 regularization, applying it with dictionary learning to develop sparse polynomial policies for deep reinforcement learning in controlling PDEs.

Contribution

It presents a novel combination of L0 regularization and dictionary learning for sparsifying models in the context of deep reinforcement learning for PDE control.

Findings

Successful sparsification of parametric models demonstrated

Development of sparse polynomial policies for PDE control

Provision of code and tutorial for reproducibility

Abstract

This document contains an educational introduction to the problem of sparsifying parametric models with L0 regularization. We utilize this approach together with dictionary learning to learn sparse polynomial policies for deep reinforcement learning to control parametric partial differential equations. The code and a tutorial are provided here: https://github.com/nicob15/Sparsifying-Parametric-Models-with-L0.