D
Derek Onken
Researcher at Emory University
Publications - 7
Citations - 139
Derek Onken is an academic researcher from Emory University. The author has contributed to research in topics: Ode & Artificial neural network. The author has an hindex of 4, co-authored 7 publications receiving 56 citations.
Papers
More filters
Posted Content
OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport.
TL;DR: The proposed OT-Flow approach tackles two critical computational challenges that limit a more widespread use of CNFs, and leverages optimal transport (OT) theory to regularize the CNF and enforce straight trajectories that are easier to integrate.
Posted Content
Discretize-Optimize vs. Optimize-Discretize for Time-Series Regression and Continuous Normalizing Flows.
Derek Onken,Lars Ruthotto +1 more
TL;DR: Disc-Opt methods can achieve similar performance as Opt-Disc at inference with drastically reduced training costs using neural ODEs for time-series regression and continuous normalizing flows (CNFs).
Posted Content
A Neural Network Approach Applied to Multi-Agent Optimal Control
TL;DR: A neural network approach for solving high-dimensional optimal control problems with obstacle and collision avoidance that fuses the Pontryagin Maximum Principle and Hamilton-Jacobi-Bellman approaches and parameterizes the value function with a neural network.
Proceedings Article
OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport
TL;DR: In this article, an optimal transport theory is used to regularize the continuous normalizing flow (CNF) and enforce straight trajectories that are easier to integrate, which can be used for density estimation and statistical inference.
Journal ArticleDOI
Cell-phone traces reveal infection-associated behavioral change.
Ymir Vigfusson,Thorgeir A. Karlsson,Derek Onken,Congzheng Song,Atli F. Einarsson,Nishant Kishore,Rebecca M. Mitchell,Ellen Brooks-Pollock,Gudrun Sigmundsdottir,Danon +9 more
TL;DR: In this paper, the authors measured behavior change reflected in mobile-phone call-detail records (CDRs), a source of passively collected real-time behavioral information, using an anonymously linked dataset of cell-phone users and their date of influenza-like illness diagnosis during the 2009 H1N1v pandemic.