Machine Learning Methods

Course guarantor:

Lecturer:

Tutor:

Supervisor:

Department of Applied Mathematics

Synopsis:

The course introduces students to machine learning methods applicable within their specializations in the follow-up Applied Informatics program. These principles and competencies are not part of the common undergraduate curriculum and are typically taught only in specializations focused on artificial intelligence. The aim is to understand the theoretical foundations and to gain practical experience in applying models suitable for regression and classification tasks within supervised learning, including kernel methods and neural networks. In unsupervised learning, students will become familiar primarily with clustering models and principal component analysis. The course also covers model evaluation techniques and fundamental methods for data preprocessing. Practical exercises involve data analysis and model implementation using the Python libraries pandas, scikit-learn, and PyTorch.

Requirements:

Syllabus of lectures:

1. Introduction and basic concepts of machine learning - supervised learning, classification and regression tasks.

2. Decision trees, nearest neighbor method, normalization.

3. Linear and ridge regression.

4. Logistic regression.

5. Ensemble methods - random forests, AdaBoost.

6. Model evaluation (regression, classification - confusion matrix and derived metrics), cross validation, feature selection.

7. Unsupervised learning - clustering (agglomerative hierarchical, k-means).

8. Support vector machine (SVM) method for classification.

9. Principal Component Analysis (PCA).

10. (2) Introduction to neural networks - loss functions, back propagation, deep learning.

11. Convolutional neural networks, recurrent neural networks.

12. Modern neural network models - autoencoders, transformers, LLM.

Syllabus of tutorials:

Study Objective:

The course introduces students to machine learning methods applicable within their specializations in the follow-up Applied Informatics program. These principles and competencies are not part of the common undergraduate curriculum and are typically taught only in specializations focused on artificial intelligence. The aim is to understand the theoretical foundations and to gain practical experience in applying models suitable for regression and classification tasks within supervised learning, including kernel methods and neural networks. In unsupervised learning, students will become familiar primarily with clustering models and principal component analysis. The course also covers model evaluation techniques and fundamental methods for data preprocessing. Practical exercises involve data analysis and model implementation using the Python libraries pandas, scikit-learn, and PyTorch.

Study materials:

1. Murphy K. P.: Machine Learning: A Probabilistic Perspective. MIT Press, 2012. ISBN 978-0-262-01802-9.

2. Goodfellow I., Bengio Y., Courville A.: Deep Learning. MIT Press, 2016. ISBN 978-0-262-03561-3.

3. Alpaydin E.: Introduction to Machine Learning. MIT Press, 2020. ISBN 978-0262043793.

4. Hastie T., Tibshirani R., Friedman J.: The Elements of Statistical Learning. Springer, 2009. ISBN 978-0-387-84857-0.

5. Deisenroth M. P.: Mathematics for Machine Learning. Cambridge University Press, 2020. ISBN 978-1108455145.

Note:

Further information:

Výukové materiály na courses

No time-table has been prepared for this course

The course is a part of the following study plans:

• Master program ANI for the phase of study without specialisation (compulsory course in the program)
• Master specialization Computer Security, in English, 2026 (VO)
• Master programme, for the phase of study without specialisation, ver. for 2026 and higher (VO)
• Master specialization Computer Systems and Networks, in English, 2026 (PS)
• Master specialization Computer Science, in English, 2026 (VO)
• Master specialization Programming Languages, in English, 2026 (VO)