COURSE INFORMATION
Course Title: INTRODUCTION TO NEURAL NETWORKS
Code Course Type Regular Semester Theory Practice Lab Credits ECTS
ECE 433 B 3 3 2 0 4 7.5
Academic staff member responsible for the design of the course syllabus (name, surname, academic title/scientific degree, email address and signature) Prof.Dr. Bekir Karlik bkarlik@epoka.edu.al
Main Course Lecturer (name, surname, academic title/scientific degree, email address and signature) and Office Hours: Prof.Dr. Bekir Karlik bkarlik@epoka.edu.al , Fridays 16-18 pm
Second Course Lecturer(s) (name, surname, academic title/scientific degree, email address and signature) and Office Hours: NA
Language: English
Compulsory/Elective: Compulsory
Study program: (the study for which this course is offered) Master of Science in Electronics and Communication Engineering
Classroom and Meeting Time: E-012 at 18:00
Teaching Assistant(s) and Office Hours: NA
Code of Ethics: Code of Ethics of EPOKA University
Regulation of EPOKA University "On Student Discipline"
Attendance Requirement:
Course Description: This course offers the basic knowledge in artificial neural networks and machine learning, spanning from the introductory concepts to well established and cardinal learning algorithms. The backpropagation learning algorithm is taught along with the delta rule for the weights’ adjustments of the neural network. Also, traditional nets are presented along with modern convolutional neural networks and deep neural networks.
Course Objectives: This course gives an overview of both the foundational ideas and the recent advances in neural net algorithms. Roughly the first 2/3 of the course focuses on supervised learning — training the network to produce a specified behavior when one has lots of labeled examples of that behavior. The last 1/3 focuses on unsupervised learning — the correct behavior isn’t specified by hand, but the goal is to discover interesting regularities in the data.
BASIC CONCEPTS OF THE COURSE
1 Introduction to ANN
2 Perceptron learning and Multi-Layer Perceptron
3 Backpropagation and Generalized Delta Rule Learning
4 ecurrent neural nets
5 Other Feedforward Networks
6 Unsupervised ANN models (Hopfield, SOM etc.)
7 Convolutional nets and Autoencoders
8 RNN, Convolutional nets and Autoencoders
9 Deep Neural Networks and Hybrid Models
10 AI ethics, Generative AI
COURSE OUTLINE
Week Topics
1 Introduction to ANN
2 Perceptron learning and Multi-Layer Perceptron
3 Backpropagation and Generalized Delta Rule Learning
4 Other Feedforward Networks
5 Probabilistic models and Bayesian neural networks
6 Unsupervised learning neural networks and SOM
7 The Other Unsupervised ANN models
8 Recurrent neural nets, Autoencoders
9 Reinforcement learning
10 Convolutional nets
11 Deep Neural Networks
12 Hybrid Models
13 ANN applications (Industrial, Medical, Business etc.)
14 Project progress report presentations
Prerequisite(s):
Textbook(s): Lecture Notes of Artificial Neural Networks, Bekir Karlik
Additional Literature: Deep Learning by Yoshua Bengio, Ian Goodfellow, and Aaron Courville
Laboratory Work: 2
Computer Usage: A programming background as Python, C/C++
Others: No
COURSE LEARNING OUTCOMES
1 Understand the learning and generalization issue in neural computation.
2 Understand the basic ideas behind most common supervised learning algorithms such as multilayer perceptron, radial- basis function networks, etc.
3 Implement common supervised learning algorithms using an existing package.
4 Apply supervised neural networks for classification, estimation, forecasting and recognition problems.
5 Present supervised learning project and write a technical report about finding results
6 Understand the basic ideas behind most common unsupervised learning algorithms such as Hopfield networks, and Kohonen self-organising maps.
7 Apply unsupervised neural networks for classification, estimation, forecasting and recognition problems.
8 Present unsupervised learning project and write a technical report about finding results
9 Understanding the basic ideas of Deep Neural Networks
10 Understanding the basic ideas of Hybrid models
COURSE CONTRIBUTION TO... PROGRAM COMPETENCIES
(Blank : no contribution, 1: least contribution ... 5: highest contribution)
No Program Competencies Cont.
Master of Science in Electronics and Communication Engineering Program
1 Engineering graduates with sufficient theoretical and practical background for a successful profession and with application skills of fundamental scientific knowledge in the engineering practice. 5
2 Engineering graduates with skills and professional background in describing, formulating, modeling and analyzing the engineering problem, with a consideration for appropriate analytical solutions in all necessary situations 5
3 Engineering graduates with the necessary technical, academic and practical knowledge and application confidence in the design and assessment of machines or mechanical systems or industrial processes with considerations of productivity, feasibility and environmental and social aspects. 4
4 Engineering graduates with the practice of selecting and using appropriate technical and engineering tools in engineering problems, and ability of effective usage of information science technologies. 5
5 Ability of designing and conducting experiments, conduction data acquisition and analysis and making conclusions. 4
6 Ability of identifying the potential resources for information or knowledge regarding a given engineering issue. 5
7 The abilities and performance to participate multi-disciplinary groups together with the effective oral and official communication skills and personal confidence. 5
8 Ability for effective oral and official communication skills in foreign language. 5
9 Engineering graduates with motivation to life-long learning and having known significance of continuous education beyond undergraduate studies for science and technology. 5
10 Engineering graduates with well-structured responsibilities in profession and ethics. 5
11 Engineering graduates who are aware of the importance of safety and healthiness in the project management, workshop environment as well as related legal issues. 5
12 Consciousness for the results and effects of engineering solutions on the society and universe, awareness for the developmental considerations with contemporary problems of humanity. 5
COURSE EVALUATION METHOD
Method Quantity Percentage
Presentation
1
20
Project
1
20
Final Exam
1
50
Attendance
10
Total Percent: 100%
ECTS (ALLOCATED BASED ON STUDENT WORKLOAD)
Activities Quantity Duration(Hours) Total Workload(Hours)
Course Duration (Including the exam week: 16x Total course hours) 16 3 48
Hours for off-the-classroom study (Pre-study, practice) 16 3 48
Mid-terms 0
Assignments 2 20 40
Final examination 1 25 25
Other 1 26.5 26.5
Total Work Load:
187.5
Total Work Load/25(h):
7.5
ECTS Credit of the Course:
7.5
CONCLUDING REMARKS BY THE COURSE LECTURER

In conclusion, this course on Artificial Neural Networks (ANNs) has provided a solid foundation in understanding the core concepts, architectures, and applications of neural networks. From the basics of perceptrons to the complexities of deep learning models, we have explored how ANNs can be applied across various domains, including computer vision, natural language processing, and more. As the field of AI continues to evolve, mastering the principles of neural networks equips us with the tools to innovate and contribute to advancements in technology. Moving forward, further exploration of more advanced techniques and real-world problem-solving will enhance our understanding and capabilities in this exciting and rapidly growing field.