Akademik Dersler - Tanıtım

Course Introduction Information


Code - Course Title	YZM3006 - Reinforcement Learning
Course Type	Required Courses
Language of Instruction	İngilizce
Laboratory + Practice	3+2
ECTS	6.0
Course Instructor(s)	DOKTOR ÖĞRETİM ÜYESİ NİHAL İNCE
Mode of Delivery	This course is offered exclusively in an in-person format.
Prerequisites	There are no prerequisites or co-requisites for this course.
Courses Recomended	There are no other recommended courses related to this one.
Required or Recommended Resources	Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction (2nd ed.). MIT Press.Szepesvári, C. (2010). Algorithms for reinforcement learning. Morgan & Claypool Publishers.Graesser, L., & Keng, W. L. (2020). Foundations of deep reinforcement learning: Theory and practice in Python. Addison-Wesley Professional.Geron, A. (2019). Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow: Concepts, tools, and techniques to build intelligent systems (2nd ed.). O'Reilly Media.Ravichandiran, S. (2018). Hands-on reinforcement learning with Python: Master reinforcement learning and deep reinforcement learning architectures using OpenAI Gym and TensorFlow. Packt Publishing.
Recommended Reading List	Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction (2nd ed.). MIT Press.Szepesvári, C. (2010). Algorithms for reinforcement learning. Morgan & Claypool Publishers.Graesser, L., & Keng, W. L. (2020). Foundations of deep reinforcement learning: Theory and practice in Python. Addison-Wesley Professional.Geron, A. (2019). Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow: Concepts, tools, and techniques to build intelligent systems (2nd ed.). O'Reilly Media.Ravichandiran, S. (2018). Hands-on reinforcement learning with Python: Master reinforcement learning and deep reinforcement learning architectures using OpenAI Gym and TensorFlow. Packt Publishing.
Assessment methods and criteria	1 midterm exam, 1 project, 1 final exam
Work Placement	This is not suitable for this class.
Sustainability Development Goals

Content

Weeks	Topics
Week - 1	What is Reinforcement Learning? Agent, Environment, Reward. Differences between ML paradigms.
Week - 2	Markov Decision Processes (MDPs). Policies and Value Functions.
Week - 3	The Bellman Expectation Equations.
Week - 4	The Bellman Optimality Equations.
Week - 5	Model-Based Solutions: Dynamic Programming - Policy Iteration.
Week - 6	Dynamic Programming - Value Iteration.
Week - 7	Model-Free Prediction: Monte Carlo (MC) Methods.
Week - 8	Model-Free Control: Temporal-Difference (TD) Learning, SARSA.
Week - 9	Q-Learning.
Week - 10	Function Approximation.
Week - 11	Deep Reinforcement Learning: The Deep Q-Network (DQN).
Week - 12	Deep Reinforcement Learning Implementation with Keras.
Week - 13	Policy Gradient Methods: The REINFORCE Algorithm.
Week - 14	Policy Gradient Methods: The Actor-Critic Architecture.

Learning Activities and Teaching Methods

Teaching Methods
Lecture
Discussion
Question & Answer
Team/Group Work
Drill - Practise
Problem Solving
Report Preparation and/or Presentation
Proje Design/Management
Competences
Productive
Rational
Questoning
Work in teams
Use time effectively
Respect differences
Eleştirel düşünebilme
Abstract analysis and synthesis
Problem solving
Information Management
To work autonomously
Elementary computing skills
Decision making
Project Design and Management

Assessment Methods

Assessment Method and Passing Requirements
	Quamtity	Percentage (%)
Toplam (%)		0