COURSE UNIT TITLE

: DEEP LEARNING

DEGREE PROGRAMMES

Description of Individual Course Units

Course Unit Code Course Unit Title Type Of Course D U L ECTS
EMT 3035 DEEP LEARNING ELECTIVE 3 0 0 5

Offered By

Econometrics

Level of Course Unit

First Cycle Programmes (Bachelor's Degree)

Course Coordinator

ASSOCIATE PROFESSOR ÖZLEM KIREN GÜRLER

Offered to

Econometrics
Econometrics (Evening)

Course Objective

The objective of this course is to introduce participants to the fundamental principles and applications of deep learning, to enable them to develop various deep learning models using the Python programming language and popular deep learning libraries (TensorFlow, Keras, PyTorch), and thus to provide them with the ability to generate deep learning solutions to real-world problems.

Learning Outcomes of the Course Unit

1   To be able to grasp the basic concepts of deep learning
2   To be able to use Python and deep learning libraries
3   To be able to develop different deep learning models
4   To be able to apply data preprocessing and model evaluation skills
5   To be able to generate deep learning solutions to real-world problems
6   To be able to develop awareness to follow developments in the field of deep learning

Mode of Delivery

Face -to- Face

Prerequisites and Co-requisites

None

Recomended Optional Programme Components

None

Course Contents

Week Subject Description
1 1. Week Introduction to Deep Learning: Course introduction and objectives. Differences between artificial intelligence, machine learning, and deep learning. History and application areas of deep learning. Basic principles of deep learning: neurons, layers, activation functions.
2 2. Week Introduction to Python and Deep Learning Libraries: Python basics (a quick review if necessary). Introduction to NumPy, Pandas, and Matplotlib libraries. TensorFlow or PyTorch installation and fundamentals. Introduction to Keras (with or without TensorFlow).
3 3. Week Artificial Neural Networks (ANN) 1: ANN architecture and working principle. Application of linear and logistic regression with deep learning. Single-layer ANNs. Activation functions (ReLU, sigmoid, tanh).
4 4. Week Artificial Neural Networks (ANN) 2: Multi-layer ANNs and the backpropagation algorithm. Loss functions and optimization algorithms (SGD, Adam). Overfitting and underfitting concepts. Regularization techniques (L1, L2).
5 5. Week Artificial Neural Networks (ANN) Applications: Solving classification problems (e.g., MNIST dataset). Solving regression problems. Model performance evaluation metrics. Hyperparameter tuning
6 6. Week Convolutional Neural Networks (CNN) 1: CNN architecture and working principle. Convolution layers and filters. Pooling layers (Max Pooling, Average Pooling). Activation functions in CNNs.
7 7. Week Convolutional Neural Networks (CNN) 2: Image classification applications. Introduction to object recognition applications. Layer organization in CNN architectures. Data augmentation techniques.
8 8. Week Convolutional Neural Networks (CNN) Applications: Developing an image classification model (e.g., CIFAR-10 dataset). Transfer learning and pre-trained models (e.g., VGG16, ResNet). Fine-tuning.
9 9. Week Recurrent Neural Networks (RNN) 1: RNN architecture and working principle. Time series data and sequential data. LSTM (Long Short-Term Memory) networks. GRU (Gated Recurrent Unit) networks.
10 10. Week Recurrent Neural Networks (RNN) 2: Introduction to natural language processing (NLP) applications. Word embeddings. Text classification. Sequence generation.
11 11. Week Recurrent Neural Networks (RNN) Applications: Developing a text classification model (e.g., IMDB dataset). Developing a time series forecasting model. Attention mechanisms in RNNs
12 12. Week Deep Learning Applications - Introduction to Project Work: Identification of student projects and formation of groups. Discussion of project topics and development of project plans. Examination of current applications in the field of deep learning.
13 13. Week Project Work: Class time allocated for students to work on their projects. Project consultation and guidance by the instructor.
14 14. Week Project Presentations and Evaluation: Student project presentations. Evaluation of projects. Overall course evaluation and closure.

Recomended or Required Reading

1. François Chollet. Deep Learning with Python. Manning Publications Co. 2017.
2. Aurelien Geron. Hands-on Machine Learning with Scikit-Learn, Keras & TensorFlow. Oreilly. 2019.
3. Hagan, M.T., Demuth, H.B. and Beale, M. Neural Network Design. 1996.

Planned Learning Activities and Teaching Methods

This course will be presented using class lectures, class discussions, overhead projections, and demonstrations

Assessment Methods

SORTING NUMBER SHORT CODE LONG CODE FORMULA
1 STT TERM WORK (SEMESTER)
2 MTE MIDTERM EXAM
3 MTEG MIDTERM GRADE STT * 0.50 +MTE * 0.50
4 FIN FINAL EXAM
5 FCG FINAL COURSE GRADE MTEG * 0.40 + FIN * 0.60
6 RST RESIT
7 FCGR FINAL COURSE GRADE (RESIT) MTEG * 0.40 + RST * 0.60


Further Notes About Assessment Methods

None

Assessment Criteria

To be announced.

Language of Instruction

Turkish

Course Policies and Rules

To be announced.

Contact Details for the Lecturer(s)

serkan.aras@deu.edu.tr

Office Hours

To be announced.

Work Placement(s)

None

Workload Calculation

Activities Number Time (hours) Total Work Load (hours)
Lectures 14 3 42
Preparations before/after weekly lectures 14 1 14
Preparation for final exam 1 14 14
Preparing assignments 1 25 25
Preparation for quiz etc. 1 15 15
Preparing presentations 1 6 6
Final 1 2 2
Midterm 1 2 2
TOTAL WORKLOAD (hours) 120

Contribution of Learning Outcomes to Programme Outcomes

PO/LOPO.1PO.2PO.3PO.4PO.5PO.6PO.7PO.8PO.9PO.10
LO.111
LO.21
LO.31
LO.411
LO.511
LO.6111

CONTACT INFORMATION
Dokuz Eylül Üniversitesi Cumhuriyet Bulvarı No: 144 35210 Alsancak / İZMİR
Phone: +90(232) 412 12 12 - Fax: +90 (232) 464 81 35

Copyright 2015 DEÜ. BİD

HOME / ABOUT US / DEGREE PROGRAMMES / GENERAL INFORMATION FOR STUDENTS / INTERNATIONAL RELATIONS