|
Lecturer(s)
|
|
|
|
Course content
|
1) Introduction to deep learning, repetition of the basics of neural networks. 2) Mathematical building blocks of neural networks, models, layers, gradient optimization methods. 3) Frameworks and libraries for working with neural networks - TensorFlow and Keras. 4) Basic methods of machine learning - learning with a teacher, learning without a teacher, self-directed and reinforced learning. 5) Image Recognition Networks - CNN. 6) Networks for text classification - LSTM. 7) Sequence data classification networks - GRU. 8) Optimal model design practices for deep learning. 9) Generative deep learning. 10) Methods for natural language processing. 11) Methods of autonomous creation of neural network models.
|
|
Learning activities and teaching methods
|
Monologic (reading, lecture, briefing)
- Semestral paper
- 44 hours per semester
- Preparation for exam
- 25 hours per semester
- Preparation for classes
- 60 hours per semester
- Class attendance
- 30 hours per semester
|
|
Learning outcomes
|
The aim of the course is to explain the known principles of machine learning, which can be applied to a wide range of problems. Students will practically try to use the given algorithms on real problems. The primary area of the subject is deep learning and its main current trends.
An overview of modern machine learning algorithms and the ability to apply them to practical problems.
|
|
Prerequisites
|
Basic knowledge of programming and methods of intelligence calculation is assumed.
|
|
Assessment methods and criteria
|
Oral examination
Create and defend the semestral project, get at least 50% of possible points from the semestral test.
|
|
Recommended literature
|
-
Chollet, F. Deeplearning v jazyku Python. Grada Publishing 2019. 328 s.. 2019. ISBN 978-80-247-3100-1.
-
Kron, J. Deep Learning Illustrated: A Visual, Interactive Guide to Artificial Intelligentce. Addison-Wesley Professional. 2019. 416 s.. 2019. ISBN 0135116694.
-
Marsland, S. Machine Learning: An algorithmic perspective (2nd edition). CRC Press. 2014. 457s.. 1997. ISBN 1466583282.
-
Mitchell, Tom M. Machine learning. The McGraw-Hill Companies, Inc. New York, 1997. 432 s.. 1997. ISBN 0070428077.
|