lidya multisite > Computer Engineering Department > CENG 567

About

CENG 567

Reinforcement Learning

Course Information
CENG 567

 

 Reinforcement Learning
 
Course Semester T+A Credits ECTS Credits
  3+0 3 9
Course Language English
Course Level Graduate
Department/Program Computer Engineering Department/Computer Engineering MS Program
Mode of delivery Face to face
Course Type Compulsory [ ] / Technical Elective [ x ]
Course Objectives This course introduces reinforcement learning, covering key techniques like multi-armed bandits, Q-learning, and deep reinforcement learning. Students will learn decision-making under uncertainty and apply advanced methods like actor-critic models to real-world problems.

 
Course Content Multi-armed bandits, epsilon greedy, upper confidence bounds, thompson sampling, contextual bandits, markov decision process, dynamic programming, policy and value iteration, monte carlo methods, temporal difference, Q-learning, deep Q-learning, actor-critic models.
Course Prerequisites  

None
Course Coordinator Asst. Prof. Dr. Osman GÖKALP
Course Lecturer(s) Asst. Prof. Dr. Osman GÖKALP
Course Assistants None
Course Internship None
   
Course Resources
Resources

Mastering Reinforcement Learning with Python, Enes Bilgin, Packt Publishing, 2020.

Reinforcement Learning, second edition: An Introduction, Richard S. Sutton and Andrew G. Barto, Bradford Books, second edition.
Planned Learning Activities and Teaching Methods
Presentations, homeworks, research.
 
Assessment Criteria ECTS Workload Calculation
In-term Studies Quantity Percentage% Activities Quantity Duration Workload (Hour)
Home works 2 20% Weekly course 14 3 42
Cases –   Outside Activities About Course (Homework, Reading, Individual studies etc.) 14 2 28
Laboratory works –  
Other activities –  
Projects –  
Quizzes –   Exams and Exam Preparations (Attendance, Presentation, Midterm

exam, Final exam, Quiz etc.)

 2 5 10
Midterm exams 1 30%
Final examination 1 50%
Total     Total Workload     80
 
Course Learning Outcomes
The students who succeeded in this course will be able to:
No Explanation
1 To understand reinforcement learning fundamentals and key concepts.
2 To apply core algorithms like Q-learning and deep Q-learning.
3 To analyze Markov decision processes and policy optimization.
4 To implement reinforcement learning solutions in real-world scenarios.
 
Weekly Course Plan
Week Topics
1 Introduction to Reinforcement Learning (RL)
2 Multi-Armed Bandits, Epsilon Greedy
3 Upper Confidence Bounds, Thompson Sampling
4 Contextual Bandits
5 Markov Decision Process
6 Dynamic Programming, Policy Iteration, Value Iteration
7 Monte Carlo Methods
8 Temporal Difference Learning, SARSA, Q-Learning
9 Deep Q-learning
10 Actor-Critic Models, A2C
11 Deep Deterministic Policiy Gradient
12 Applications of RL – I
13 Applications of RL – II
14 Overview of the Course and Feedbacks

Instructor(s)

Assistant Professor
Osman Gökalp
Other MS Courses