Overview
Explore cutting-edge techniques for autonomous navigation in urban environments without relying on traditional maps. Delve into reinforcement learning concepts, including value functions and Q-value functions, as applied to city navigation. Examine the courier task problem, analyze various actions and environments, and understand the architecture of a City Navigation Agent. Learn about training methodologies, the actor-critic approach, and multi-site experiments. Gain insights into abolition analysis and goal description techniques for improving navigation performance in complex urban settings.
Syllabus
Introduction
What is Reinforcement Learning
Value Function and QValue Function
Problem Statement
Actions
Environments
Courier task
Architecture
City Navigation Agent
Training
Act of Creating
Actor Critic
Results
Multisite Experiments
Abolition Analysis
Goal Description
Taught by
UCF CRCV