CM30359: Reinforcement learning
[Page last updated: 26 October 2022]
| Academic Year: | 2022/23 |
| Owning Department/School: | Department of Computer Science |
| Credits: | 6 [equivalent to 12 CATS credits] |
| Notional Study Hours: | 120 |
| Level: | Honours (FHEQ level 6) |
| Period: |
|
| Assessment Summary: | CW 100% |
| Assessment Detail: |
|
| Supplementary Assessment: |
|
| Requisites: |
Before taking this module you must take CM10310 OR take CM20220 OR take CM20252
In taking this module you cannot take CM50270 |
| Learning Outcomes: | On completion of the unit, the students will be able to:
1. formulate reinforcement learning problems by defining a state space, an action space, and a reward function, appropriate for the context, 2. apply a range of solution methods to reinforcement learning problems, 3. appreciate the difficulties encountered in solving large, complex reinforcement learning problems in practice. |
| Aims: | To explore reinforcement learning as an approach to artificial intelligence; to understand how reinforcement learning differs from other approaches to machine learning such as supervised and unsupervised learning; to learn how to formulate and solve reinforcement learning problems, and to appreciate the difficultes involved in solving large, complex reinforcement learning problems in practice. |
| Skills: | Intellectual skills:
* Conceptual understanding of sequential decision making under uncertainty (T, F, A) Practical skills: * Programming reinforcement learning algorithms (T, F, A) * Use of software libraries for reinforcement learning (T, F, A) Transferable skills: * Technical report writing (F, A) |
| Content: | Topics covered normally include: dynamic programming, Monte Carlo methods, temporal-difference algorithms (e.g., Q-learning), integration of planning and learning, value function approximation (e.g., with deep neural networks), policy-gradient methods, application areas, and an introduction to active areas of research (e.g., hierarchical reinforcement learning, intrinsically-motivated reinforcement learning systems). |
| Programme availability: |
CM30359 is Optional on the following programmes:Department of Computer Science
|
Notes:
|