COMP6008

COMP6008: Reinforcement Learning

Student reviews for COMP6008 at Curtin University. See ratings, workload, grades, and honest feedback from Curtin students.

View in Curtin Handbook

COMP6008 student reviews

No reviews of COMP6008 yet. Be the first: the workload it really took, how it was assessed, and whether you would take it again.

Which units require COMP6008?

No other unit in the catalogue lists COMP6008 as a prerequisite, so it is not a gateway to anything further.

What is COMP6008 about?

This unit will provide introduction to reinforcement learning and relevant practical background to solve challenging real-world problems. It covers both model-free and model-based approaches. Students will study tabular reinforcement learning methods such as multi-armed bandits, dynamic programming, temporal-difference learning including Q-learning and SARSA. Different policy search techniques, such as policy gradient methods, are discussed. Finally, the unit also covers advanced reinforcement learning topics related to deep learning and practical considerations.