Information Package / Course Catalogue
| Computer Science and Engineering (English) | |||||
| PhD | Length of the Programme: 4 | Number of Credits: 240 | TR-NQF-HE: Level 8 | QF-EHEA: Third Cycle | EQF: Level 8 |
Ders Genel Tanıtım Bilgileri
| School/Faculty/Institute | Graduate School | ||||||
| Course Code | CSE 618 | ||||||
| Course Title in English | Advanced Artificial Neural Networks | ||||||
| Course Title in Turkish | İleri Yapay Sinir Ağları | ||||||
| Language of Instruction | EN | ||||||
| Type of Course | Flipped Classroom | ||||||
| Level of Course | Advanced | ||||||
| Semester | Fall | ||||||
| Contact Hours per Week |
|
||||||
| Estimated Student Workload | 187 hours per semester | ||||||
| Number of Credits | 7.5 ECTS | ||||||
| Grading Mode | Standard Letter Grade | ||||||
| Pre-requisites | None | ||||||
| Expected Prior Knowledge | Artificial Intelligence; Neural Networks | ||||||
| Co-requisites | None | ||||||
| Registration Restrictions | Only Doctorate Students | ||||||
| Overall Educational Objective | To become familiar with a comprehensive understanding of foundational artificial neural network architectures and algorithms, and to apply these networks effectively to solve real-world problems. | ||||||
| Course Description | Advanced Artificial Neural Networks course delves into the core architectures and algorithms of artificial neural networks. Students will be introduced to topics such as perceptrons, linear regression, backpropagation, support vector machines, and self-organizing maps. The course aims to equip learners with the skills to employ these networks in tackling real-world challenges. | ||||||
| Course Description in Turkish | Gelişmiş Yapay Sinir Ağları dersi, yapay sinir ağlarının temel mimarileri ve algoritmalarını detaylı bir şekilde ele almaktadır. Öğrencilere perceptronlar, doğrusal regresyon, geri yayılım, destek vektör makinaları ve kendi kendini düzenleyen haritalar gibi konularda bilgi verilmektedir. Ders, bu ağları gerçek dünya sorunlarının çözümünde nasıl kullanacakları konusunda öğrencilere beceriler kazandırmayı hedeflemektedir. |
Course Learning Outcomes and CompetencesUpon successful completion of the course, the learner is expected to be able to:1) Explain and derive mathematical expressions related to neural network models; 2) Implement and optimize neural network training algorithms from scratch; 3) Formulate real-world problems in terms of neural network tasks and select appropriate models for solutions 4) Differentiate between various advanced models and decide when to deploy them; 5) Use neural network libraries and platforms, such as TensorFlow or PyTorch, to build and test models 6) Use evaluation metrics, such as accuracy, precision, recall, and F1 score, and apply cross-validation methods to assess model performance 7) Analyze potential biases in datasets, understand the societal ramifications of model decisions, and advocate for responsible AI use. |
| Program Learning Outcomes/Course Learning Outcomes | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
|---|
Relation to Program Outcomes and Competences
| N None | S Supportive | H Highly Related |
| Program Outcomes and Competences | Level | Assessed by |
| Prepared by and Date | , |
| Course Coordinator | TUNA ÇAKAR |
| Semester | Fall |
| Name of Instructor | Asst. Prof. Dr. TUNA ÇAKAR |
Course Contents
| Week | Subject |
| 1) | Introduction to Artificial Neural Networks Overview of neural networks and their historical evolution. |
| 2) | Perceptrons Single-layer perceptrons, their structure, and the perceptron learning rule. |
| 3) | Linear Regression and Least Mean Squares Introduction to regression, linear models, and optimization through the least mean squares method. |
| 4) | Multi-layer Perceptrons (MLPs) Structure of MLPs, activation functions, and introduction to feedforward mechanisms. |
| 5) | Backpropagation Algorithm Understanding the error propagation, gradient descent, and weight updates in multi-layer networks. |
| 6) | Support Vector Machines (SVM) Introduction to SVM, its advantages, and kernel methods. |
| 7) | Radial Basis Function (RBF) Networks Structure of RBF, its advantages, and applications. |
| 8) | Self-organizing Maps (SOM) Unsupervised learning, Kohonen networks, and feature mapping. |
| 9) | Recurrent Neural Networks (RNNs) Introduction to sequences, basic RNN structures, and challenges like vanishing and exploding gradients. |
| 10) | Optimization Techniques in Neural Networks Momentum, RMSprop, Adam optimizer, and learning rate scheduling. |
| 11) | Regularization and Overfitting in Neural Networks Techniques like dropout, early stopping, and L1 & L2 regularization. |
| 12) | Advanced Architectures Overview Brief exploration of architectures like Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTM). |
| 13) | Ethical Implications and Fairness in Neural Networks Understanding biases, ethical considerations, and ensuring fairness in model predictions. |
| 14) | Course Review and Emerging Trends Recap of the course topics, and a glimpse into current research and emerging trends in neural networks. |
| 15) | Final Examination Period |
| 16) | Final Examination Period |
| Required/Recommended Readings | None | ||||||||||||
| Teaching Methods | Lecturing and exercises in the classroom with computers. In-class exercises and 3 Projects will be carried out by students | ||||||||||||
| Homework and Projects | In-class exercises, 2 Projects, Term project | ||||||||||||
| Laboratory Work | Programming exercises | ||||||||||||
| Computer Use | For programming | ||||||||||||
| Other Activities | None | ||||||||||||
| Assessment Methods |
|
||||||||||||
| Course Administration |
cakart@mef.edu.tr 02123953600 Instructor’s office and phone number, office hours, email address: To be announced -Office: 5th Floor, #18; Phone number: 0 212 395 37 50; Email: cakart@mef.edu.tr Rules for attendance: Minimum of 70% attendance required. Missing a quiz: Provided that proper documents of excuse are presented, each missed quiz by the student will be given a grade which is equal to the average of all of the other quizzes. No make-up will be given. Missing a midterm: Provided that proper documents of excuse described in the MEF University graduate regulations are presented, a make-up will be given Missing a final: Provided that proper documents of excuse described in the MEF University graduate regulations are presented, a make-up will be given Statement on academic dishonesty and plagiarism: Law on Higher Education Article 54. |
||||||||||||
ECTS Student Workload Estimation
| Activity | No/Weeks | Hours | Calculation | ||||
| No/Weeks per Semester | Preparing for the Activity | Spent in the Activity Itself | Completing the Activity Requirements | ||||
| Course Hours | 14 | 1 | 3 | 1 | 70 | ||
| Project | 2 | 18 | 2 | 40 | |||
| Quiz(zes) | 3 | 4 | 1 | 15 | |||
| Midterm(s) | 2 | 15 | 2 | 34 | |||
| Paper Submission | 1 | 25 | 3 | 28 | |||
| Total Workload | 187 | ||||||
| Total Workload/25 | 7.5 | ||||||
| ECTS | 7.5 | ||||||