Master
2023/2024
Deep Learning
Type:
Compulsory course (Machine Learning and Data Analysis)
Area of studies:
Applied Mathematics and Informatics
Delivered by:
Department of Informatics
When:
1 year, 3, 4 module
Mode of studies:
offline
Open to:
students of one campus
Instructors:
Sergei Koltsov
Master’s programme:
Machine Learning and Data Analysis
Language:
English
ECTS credits:
6
Contact hours:
84
Course Syllabus
Abstract
The discipline of choice. Deep learning is a popular area that uses neural networks of complex architecture. Such systems give better results in areas such as image and video processing, sound and text. The course will cover the main types of architectures and the principles of operation and training of deep neural networks and conduct practice in the above areas of application. To master the discipline, the student must have knowledge in the field of machine learning and optimization methods.
Learning Objectives
- The objectives of mastering the discipline "Deep Learning" are the formation of students' theoretical knowledge and practical skills on the basics of building large neural networks for deep learning.
Expected Learning Outcomes
- He knows the general, general scientific and business vocabulary used in the field of deep learning. It receives from articles (including in English) information about the structure of the neural network and the features used to solve a specific problem. Describes, presents and analyzes the results of applying deep learning methods to solve applied problems.
- He knows the main types of tasks solved using deep learning. Develops architecture, implements, trains and optimizes the parameters of neural networks. It solves applied problems from various fields using deep learning.
- Knows optimization algorithms for deep neural networks based on various variations of gradient descent. Configures such algorithms based on the conditions of a specific task
- Knows the basic architecture of neural networks used to classify images. Modifies layers and various parameters to solve applied problems. It uses networks to solve the problems of image classification, image segmentation and video stream.
Course Contents
- Optimization and Regularization Algorithms
- Image Processing and Analysis
- Natural Language Processing, Competitive and Generative Neural Networks
- Hyperparameter Optimization, Reinforcement Learning
Assessment Elements
- Homework №2Homework №2 is given to students in one version and consists of 3 tasks. Homework completion time is 4 weeks. The form of presentation of homework is an algorithm implemented in any programming language.
- Homework №1Homework №1 is given to students in one version and consists of 3 tasks. Homework completion time is 4 weeks. The form of presentation of homework is an algorithm implemented in any programming language.
- Homework №3Homework №3 is given to students in one version and consists of 3 tasks. Homework completion time is 4 weeks. The form of presentation of homework is an algorithm implemented in any programming language.
- Exam №1The oral examination is conducted in the form of answers to the questions of the examination ticket. The exam ticket contains two questions from the list of questions for the exam. Additional questions are possible if the examinee did not answer the ticket questions in sufficient detail. 2.5 hours are allotted for the preparation of the answer.
- Exam №2The oral examination is conducted in the form of answers to the questions of the examination ticket. The exam ticket contains two questions from the list of questions for the exam. Additional questions are possible if the examinee did not answer the ticket questions in sufficient detail. 2.5 hours are allotted for the preparation of the answer.
Interim Assessment
- 2023/2024 4th module0.2 * Exam №1 + 0.2 * Exam №2 + 0.2 * Homework №1 + 0.2 * Homework №2 + 0.2 * Homework №3
Bibliography
Recommended Core Bibliography
- Iba, H. (2018). Evolutionary Approach to Machine Learning and Deep Neural Networks : Neuro-Evolution and Gene Regulatory Networks. Singapore: Springer. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=1833749
Recommended Additional Bibliography
- Taweh Beysolow II. (2017). Introduction to Deep Learning Using R. Springer. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsrep&AN=edsrep.b.spr.sprbok.978.1.4842.2734.3