Browsing Theses by Supervisor "Liu, Jun"
Now showing items 1-6 of 6
-
Bidirectional TopK Sparsification for Distributed Learning
(University of Waterloo, 2022-05-27)Training large neural networks requires a large amount of time. To speed up the process, distributed training is often used. One of the largest bottlenecks in distributed training is communicating gradients across different ... -
Control of Non-deterministic Transition Systems for Linear Temporal Logic Specifications
(University of Waterloo, 2021-02-01)Consider the formal synthesis problem in the continuous dynamical systems. A systematic approach is the abstraction-based method: constructing an abstraction of the original continuous system in the discrete space, and ... -
Invariant manifold theory for impulsive functional differential equations with applications
(University of Waterloo, 2019-07-03)The primary contribution of this thesis is a development of invariant manifold theory for impulsive functional differential equations. We begin with an in-depth analysis of linear systems, immersed in a nonautonomous ... -
Kinodynamic Planning with μ-Calculus Specifications
(University of Waterloo, 2018-09-19)Motion planning problems involve determining appropriate control inputs to guide a system towards a desired endpoint. Sampling-based motion planning was developed as a technique for discretizing the state space of systems ... -
Model-based Reinforcement Learning of Nonlinear Dynamical Systems
(University of Waterloo, 2022-01-25)Model-based Reinforcement Learning (MBRL) techniques accelerate the learning task by employing a transition model to make predictions. In this dissertation, we present novel techniques for online learning of unknown dynamics ... -
Robustly Complete Temporal Logic Control Synthesis for Nonlinear Systems
(University of Waterloo, 2019-12-18)Modern systems such as spacecrafts and autonomous vehicles are complex yet safety-critical, and therefore the control methods that can deal with different dynamics and constraints while being provably correct are sought ...