Continuous Space

Continuous space research focuses on developing methods to efficiently model and analyze systems with infinitely many possible states and actions, a challenge arising in diverse fields from robotics and machine learning to game theory and physics. Current research emphasizes developing efficient algorithms and model architectures, such as neural networks and hierarchical transformers, to handle the computational complexity of continuous spaces, often incorporating techniques like sampling-based planning, gradient flows, and optimal transport. These advancements are crucial for improving the performance and scalability of AI systems, enabling more realistic simulations, and facilitating the solution of complex optimization and control problems in various scientific and engineering domains.

Papers

December 23, 2022

Online Planning for Constrained POMDPs with Continuous Spaces through Dual Ascent
Arec Jamgochian, Anthony Corso, Mykel J. Kochenderfer
Markov Decision Process Budget Constraint Continuous Space Spatial Planning Continuous POMDP

December 16, 2022

Learning on Persistence Diagrams as Radon Measures
Alex Elchesen, Iryna Hartsock, Jose A. Perea, Tatum Rask
LeArning Abstract Supervised Learning Novel Descriptor Continuous Space Persistence Diagram Indoor Radon

November 4, 2022

Modeling Temporal Data as Continuous Functions with Stochastic Process Diffusion
Marin Biloš, Kashif Rasul, Anderson Schneider, Yuriy Nevmyvaka, Stephan Günnemann
Diffusion Model Probabilistic Forecast Temporal Modeling Continuous Space Sampled Time Series Stochastic Diffusion

September 29, 2022

Ensemble Reinforcement Learning in Continuous Spaces -- A Hierarchical Multi-Step Approach for Policy Training
Gang Chen, Victoria Huang
Reinforcement Learning Deep Reinforcement Learning Actor Critic Multi Step Based Policy Continuous Space Ensemble Algorithm Ensemble Deep Reinforcement Learning

August 28, 2022

Neural Network Approximation of Continuous Functions in High Dimensions with Applications to Inverse Problems
Santhosh Karnik, Rongrong Wang, Mark Iwen
Neural Network Financial Application Inverse Problem High Dimension Low Complexity High Dimensional Space Continuous Space Neural Network Approximation

June 22, 2022

Learning Optimal Treatment Strategies for Sepsis Using Offline Reinforcement Learning in Continuous Space
Zeyu Wang, Huiying Zhao, Peng Ren, Yuxi Zhou, Ming Sheng
Deep Reinforcement Learning Offline Reinforcement Learning Real Time Continuous Space Dynamic Treatment Regime Standard Sepsis Optimal Treatment Medical Decision

May 18, 2022

Neural ODE Control for Trajectory Approximation of Continuity Equation
Karthik Elamvazhuthi, Bahman Gharesifard, Andrea Bertozzi, Stanley Osher
Neural ODE Neural Ordinary Differential Equation Controllability Condition Continuous Space Structural Controllability

March 24, 2022

Kullback-Leibler control for discrete-time nonlinear systems on continuous spaces
Kaito Ito, Kenji Kashima
Optimal Control Nonlinear Dynamic Discrete Time Kullback Leibler Divergence Controllability Condition Continuous Space

January 24, 2022

CTRMs: Learning to Construct Cooperative Timed Roadmaps for Multi-agent Path Planning in Continuous Spaces
Keisuke Okumura, Ryo Yonetani, Mai Nishimura, Asako Kanezaki
Single Agent Continuous Space Multi Agent Pathfinding Probabilistic Roadmap Multi Agent Path Planning

November 18, 2021

Gradient flows on graphons: existence, convergence, continuity equations
Sewoong Oh, Soumik Pal, Raghav Somani, Raghavendra Tripathi
Early Stage Convergence Natural Gradient Gradient Flow Wasserstein Gradient Continuous Space Gradient Scaling

Continuous Space

Papers

Online Planning for Constrained POMDPs with Continuous Spaces through Dual Ascent

Learning on Persistence Diagrams as Radon Measures

Modeling Temporal Data as Continuous Functions with Stochastic Process Diffusion

Ensemble Reinforcement Learning in Continuous Spaces -- A Hierarchical Multi-Step Approach for Policy Training

Neural Network Approximation of Continuous Functions in High Dimensions with Applications to Inverse Problems

Learning Optimal Treatment Strategies for Sepsis Using Offline Reinforcement Learning in Continuous Space

Neural ODE Control for Trajectory Approximation of Continuity Equation

Kullback-Leibler control for discrete-time nonlinear systems on continuous spaces

CTRMs: Learning to Construct Cooperative Timed Roadmaps for Multi-agent Path Planning in Continuous Spaces

Gradient flows on graphons: existence, convergence, continuity equations