On the Dynamics of Learning Time-Aware Behavior with Recurrent Neural Networks

no code implementations • 12 Jun 2023 • Peter DelMastro, Rushiv Arora, Edward Rietman, Hava T. Siegelmann

In this way, we demonstrate how dynamical systems theory can provide insights into not only the learned representations of these models, but also the dynamics of the learning process itself.

A Search and Detection Autonomous Drone System: from Design to Implementation

no code implementations • 29 Nov 2022 • Mohammadjavad Khosravi, Rushiv Arora, Saeede Enayati, Hossein Pishro-Nik

In order to obtain the shortest path to the target, a general framework is provided to model the problem of target detection when the target's location is probabilistically known.

Bayesian Inference Management

Locally Constrained Representations in Reinforcement Learning

no code implementations • 20 Sep 2022 • Somjit Nath, Rushiv Arora, Samira Ebrahimi Kahou

This encourages the representations to be driven not only by the value/policy learning but also by an additional loss that constrains the representations from over-fitting to the value loss.

Continuous Control reinforcement-learning +2

Model-Based Reinforcement Learning with SINDy

no code implementations • 30 Aug 2022 • Rushiv Arora, Bruno Castro da Silva, Eliot Moss

We found that an optimal policy trained on the discovered dynamics of the underlying system can generalize well.

Model-based Reinforcement Learning reinforcement-learning +1