Učni načrt predmeta

Predmet:
Avtomatizirano strojno učenje in optimizacija
Course:
Automated Machine Learning and Optimization
Študijski program in stopnja /
Study programme and level		 Študijska smer /
Study field		 Letnik /
Academic year		 Semester /
Semester
Informacijske in komunikacijske tehnologije, 3. stopnja 1 1
Information and Communication Technologies, 3rd cycle 1 1
Vrsta predmeta / Course type
Izbirni / Elective
Univerzitetna koda predmeta / University course code:
Predavanja
Lectures		 Seminar
Seminar		 Vaje
Tutorial		 Klinične vaje
work		 Druge oblike
študija		 Samost. delo
Individ. work		 ECTS
15 15 15 105 5

*Navedena porazdelitev ur velja, če je vpisanih vsaj 15 študentov. Drugače se obseg izvedbe kontaktnih ur sorazmerno zmanjša in prenese v samostojno delo. / This distribution of hours is valid if at least 15 students are enrolled. Otherwise the contact hours are linearly reduced and transfered to individual work.

Nosilec predmeta / Course leader:
doc. dr. Tome Eftimov
Sodelavci / Lecturers:
Jeziki / Languages:
Predavanja / Lectures:
slovenščina, angleščina / Slovenian, English
Vaje / Tutorial:
Pogoji za vključitev v delo oz. za opravljanje študijskih obveznosti:
Prerequisites:

Zaključen študij druge stopnjes področja informacijskih ali komunikacijskih tehnologij ali zaključen študij druge stopnje na drugih področjih z znanjem osnov s področja predmeta. Potrebna so tudi osnovna znanja matematike, računalništva in informatike.

Completed second-cycle studies in information or communication technologies or completed second-cycle studies in other fields with knowledge of fundamentals in the field of this course. Basic knowledge of mathematics, computer science and informatics is also requested.

Vsebina:
Content (Syllabus outline):

1. Strojno učenje
2. Meta-učenje
3. Bayesianova optimizacija
4. Izbor reprezentativnih učnih podatkov
5. Izbor algoritmov
6. Konfiguracija algoritma
7. Ocenjevanje robustne statistične uspešnosti
8. Primeri v metaheuristikah
9. Primeri v strojnem učenju

1. Machine learning
2. Meta-learning
3. Bayesian optimization
4. Selection of representative learning data
5. Algorithm selection
6. Algorithm configuration
7. Robust statistical performance assestment
8. Examples in meta-herustics
9. Examples in ML

Temeljna literatura in viri / Readings:

Izbrana poglavja iz naslednjih knjig: / Selected chapters from the following books:
- Zou, L. (2022). Meta-learning: Theory, algorithms and applications. Elsevier.
- Eftimov, T., & Korošec, P. (2022). Deep Statistical Comparison for Meta-heuristic Stochastic Optimization Algorithms. Springer Nature.
- Hutter, F., Kotthoff, L., & Vanschoren, J. (2019). Automated machine learning: methods, systems, challenges (p. 219). Springer Nature.
- Das, S., & Cakmak, U. M. (2018). Hands-On Automated Machine Learning: A beginner's guide to building automated machine learning systems using AutoML and Python. Packt Publishing Ltd.
- Pillay, N., & Qu, R. (Eds.). (2021). Automated Design of Machine Learning and Search Algorithms. Berlin/Heidelberg, Germany: Springer.

Cilji in kompetence:
Objectives and competences:

Cilj predmeta je seznaniti študenta s področjem avtmatiziranega strojnega učenja in optimizacije.

Kompetence študenta z uspešno zaključenim predmetom bodo vključevale razumevanje osnovnih pojmov iz področja, poznavanje sodobnih metod in znanje o primerih uporabe le-teh na vse bolj pomembnem znanstvenem področju.

The goal of the course is to familiarize the student with the automated machine learning and optimization.

The competencies of the students completing this course successfully would include understanding of basic concepts from both areas, familiarity with state-of-the art methods, and knowledge of examples applications from the advancing scientific field.

Predvideni študijski rezultati:
Intendeded learning outcomes:

Študenti bodo z uspešno opravljenimi obveznostmi tega predmeta pridobili:
- Pregled osnovnih konceptov strojnega učenja
- Razumevanje konceptov meta-učenja, vključno z analizo krajine
- Razumevanje metod za izbiro reprezentativnih učnih podatkov
- Razumevanje metod za izbiro algoritmov in konfiguracijo algoritmov
- Sposobnost uporabe obstoječih metod pri učnih nalogah strojnega učenja
- Sposobnost uporabe obstoječih metod pri nalogah optimizacije

Students successfully completing this course will acquire:
- Overview of basic concepts in machine learning
- Understanding concepts of meta-learning including landscape analysis
- Understanding methods for selecting representative learning data
- Understanding methods for algorithm selection and algorithm configuration
- The ability to apply existing methods in ML learning tasks
- The ability to apply existing methods in optimization tasks

Metode poučevanja in učenja:
Learning and teaching methods:

Predavanja, konzultacije, druge metode

Lectures, consultations, other methods

Načini ocenjevanja:
Delež v % / Weight in %
Assesment:
Ustni izpit
50 %
Oral Exam
Seminarska naloga
50 %
Seminar Work
Reference nosilca / Lecturer's references:
1. Gjorgjevikj, A., Nikolikj, A., Seljak, B. K., & Eftimov, T. (2025). User-defined trade-offs in LLM benchmarking: balancing accuracy, scale, and sustainability. Knowledge-Based Systems, 114405.
2. Nikolikj, A., Munoz, M. A., & Eftimov, T. (2025). Benchmarking footprints of continuous black-box optimization algorithms: Explainable insights into algorithm success and failure. Swarm and Evolutionary Computation, 94, 101895.
3. Cenikj, G., Petelin, G., & Eftimov, T. (2024). A cross-benchmark examination of feature-based algorithm selector generalization in single-objective numerical optimization. Swarm and Evolutionary Computation, 87, 101534.
4. Kostovska, A., Cenikj, G., Vermetten, D., Jankovic, A., Nikolikj, A., Skvorc, U., ... & Eftimov, T. (2023, September). PS-AAS: Portfolio Selection for Automated Algorithm Selection in Black-Box Optimization. In International Conference on Automated Machine Learning (AutoML 2023).
5. Kostovska, A., Vermetten, D., Džeroski, S., Panov, P., Eftimov, T., & Doerr, C. (2023, April). Using Knowledge Graphs for Performance Prediction of Modular Optimization Algorithms. In International Conference on the Applications of Evolutionary Computation (Part of EvoStar) (pp. 253-268). Cham: Springer Nature Switzerland