Reinforcement Learning pro ovládání robotů

Abstract

The goal of this master thesis is to design and test methods for teaching robots to perform simple tasks within a robotic simulator. Teaching is realized by using reinforcement learning algorithms. The work describes machine learning methods and algorithms for reinforcement learning. The implementation part follows, which describes the design and implementation of the solution. The result of the work is teaching the robot to perform a task, the aim of which is to move the cube to the selected location.