| CODE | ARI3212 | ||||||||
| TITLE | Advanced Reinforcement Learning | ||||||||
| UM LEVEL | 03 - Years 2, 3, 4 in Modular Undergraduate Course | ||||||||
| MQF LEVEL | 6 | ||||||||
| ECTS CREDITS | 5 | ||||||||
| DEPARTMENT | Artificial Intelligence | ||||||||
| DESCRIPTION | Reinforcement Learning is one of the major AI paradigms which in the last years lead to major milestones in the advancement of AI. From the ability to beat the world champion in the game of Go, or winning professional players in Dota, the immense potential of Reinforcement Learning is also being applied to various industry applications such as manufacturing plant control, self driving cars, algorithmic trading and portfolio management, route management and many more. The study-unit brings together advanced techniques from Reinforcement Learning in conjunction with Deep Neural Networks and Evolutionary Algorithms, the latter two being core components of Advanced Reinforcement Learning models. Topics covered in this unit include: - Mathematical framework using Markov Decision Processes - Model-Free Prediction and Control - Value Function Approximation - Neural Networks and Deep Neural Networks - Deep Reinforcement Learning - Policy Gradient Methods - Evolutionary Algorithms - Evolutionary Reinforcement Learning. Study-unit Aims: - Teach the core mathematical and algorithmic components that students would require to understand the latest advancements in Reinforcement Learning; - Cover deep and evolutionary learning techniques as applied in a Reinforcement Learning framework; - Demonstrate how complex problems can be framed as a Reinforcement Learning problem. Learning Outcomes: 1. Knowledge & Understanding By the end of the study-unit the student will be able to: - Define a mathematical and algorithmic framework to control and predict sequential decision processes; - Frame complex problems in the context of a Reinforcement Learning technique; - Identify the strengths and weaknesses, and applicability, of the different techniques. 2. Skills By the end of the study-unit the student will be able to: - Determine how to address complex problems that are sequential in nature by automating sequential decisions with the aim of maximizing long term rewards; - Implement a solution using a Reinforcement Learning approach. Main Text/s and any supplementary readings: Main text: - Richard S Sutton, Andrew G Barto “Reinforcement Learning: An Introduction”, A Bradford Book, 5 Hayward Street, Cambridge, MA, United States, 2018. Supplementary readings: - Miguel Morales “Grokking Deep Reinforcement Learning”, Manning Publications, 2020. - Maxim Lapan “Deep Reinforcement Learning Hands-On”, Pact Publishing, 2nd Edition, 2020. |
||||||||
| ADDITIONAL NOTES | Pre-Requisites: Reinforcement Learning (being proposed at Level 2) Pre-Requisite Study-Unit: ARI2204 |
||||||||
| STUDY-UNIT TYPE | Lecture, Independent Study and Project | ||||||||
| METHOD OF ASSESSMENT |
|
||||||||
| LECTURER/S | Vincent Vella |
||||||||
|
The University makes every effort to ensure that the published Courses Plans, Programmes of Study and Study-Unit information are complete and up-to-date at the time of publication. The University reserves the right to make changes in case errors are detected after publication.
The availability of optional units may be subject to timetabling constraints. Units not attracting a sufficient number of registrations may be withdrawn without notice. It should be noted that all the information in the description above applies to study-units available during the academic year 2024/5. It may be subject to change in subsequent years. |
|||||||||