This module aims to teach students the theory and implementation of reinforcement learning. Topics include:  

• Supervised learning: the backpropagation algorithm (as prerequisite for Deep reinforcement learning).
• Reinforcement Learning: Temporal Difference Learning (Q learning, SARSA), Deep Reinforcement Learning, Advanced Topics.

As well as the material taught in class, students are expected to self-study relevant books and research articles and produce reports in research article styles.

• Assignment (threshold) [Pass/Fail]
• Formal examination (grading) [60%]

Threshold assessment must be passed in order to pass the module - not passing will result in an overall module mark of zero.

• Blackboard
• Unconfirmed practical marks when available
• Exam Papers, past 2 years (where applicable)

• demonstrate understanding of the working principles and mathematical theory of reinforcement learning algorithms.
• demonstrate the ability to implement these methods from the first principles.
• demonstrate the ability to apply these methods to sample problems.

• Supervised learning: the backpropagation algorithm (as prerequisite for Deep reinforcement Learning).
• Reinforcement Learning: (Temporal Difference Learning: Q learning, SARSA, Deep Reinforcement Learning, Advanced topics. 

COMU101, COMU103, COMU06, COMU05, COMU117, COMU109, COMU118. Students from schools other than Computer Science will need to demonstrate an excellent understanding of programming (Python or Matlab) and mathematics. A level math is compulsory.

Optional modules within the school have limited capacity. We will always try to accommodate all students but cannot guarantee a place.