Subject » BMEVITMMSMB001-00
Advanced Machine Learning
Haladó gépi tanulási algoritmusok
A tantárgyleírás hatályossága
Hatályosság kezdete:
—
Hatályosság vége:
—
| Subject name (Hungarian, English) |
Haladó gépi tanulási algoritmusok
Advanced Machine Learning
|
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Subject code | BMEVITMMSMB001-00 | ||||||||||||
| Subject type | — | ||||||||||||
| Training Level | — | ||||||||||||
| Course types and hours (weekly/semester) |
|
||||||||||||
| Assessment type | vizsga | ||||||||||||
| Credits | 5 | ||||||||||||
| Subject coordinator |
DR. Gyires-Tóth Bálint Pál
position: egyetemi docens
contact:
gyires-toth.balint@vik.bme.hu
|
||||||||||||
| Responsible department |
—
|
||||||||||||
| Faculty | |||||||||||||
| Subject website | — | ||||||||||||
| Primary curriculum type | — | ||||||||||||
| Direct prerequisites – Strong prerequisite | none | ||||||||||||
| Direct prerequisites – Weak prerequisite | none | ||||||||||||
| Direct prerequisites – Parallel prerequisite | none | ||||||||||||
| Direct prerequisites – Milestone prerequisite | none | ||||||||||||
| Direct prerequisites – Exclusion | none |
Objectives
Programme
MLOps/AIOps: The machine learning/deep learning lifecycle, its characteristic challenges, and the levels of MLOps / AIOps maturity. Data and training pipelines, experiment tracking, reproducibility, version control, and continuous integration and delivery (CI/CD) adapted to machine learning. Deployment and serving of models, the fundamentals of accelerator hardware and resource sizing, scaling of inference, as well as production monitoring and the handling of model/data drift. LLMOps special aspects.
Deep learning-based time series modeling: The characteristics of temporal data and leakage-free, time-order-aware evaluation. Baseline models. Recurrent, convolutional, and transformer-based predictive architectures, as well as time series foundation models and zero-/few-shot methods.
Foundation models, transfer learning, and efficiency: The mechanisms behind the success of large language models (LLMs) and the scaling laws; the role of transfer learning in few-shot and zero-shot tasks. Parameter-efficient fine-tuning (PEFT) and the distributed scaling of large-model training, as well as model compression (for example quantization and distillation) and efficient inference for more efficient operation.
Methods for interpreting transformer representations and their use: neural probing, mechanistic analysis, and functional network (circuit) analysis, for the purpose of model tracking, steering, and direct modification.
Advanced probabilistic inference and sampling: Hamiltonian Monte Carlo and Langevin dynamics, with particular attention to stochastic gradient methods such as SGLD and their Riemannian variants. The theoretical and practical connections of score-based diffusion models. Applications in Bayesian deep learning and posterior approximation.
Causal discovery in continuous domains: gradient-based causal structure learning techniques, including NOTEARS and its neural extensions such as DAG-GNN and iNOTEARS. Functional causal models, the identifiability of causal structures in nonlinear systems, and the differences between interventional and purely observational causal learning.
Bayesian causal structure learning: Introducing the Bayesian approach to learning causal DAGs, focusing on Bayesian model averaging, BayesDAG, and commonly used scoring metrics such as BGe and BDeu. MCMC sampling in DAG spaces, structure prior probabilities, and non-parametric Bayesian models.
Advanced and personalized federated learning: addresses the optimization challenges arising from data heterogeneity across clients. Approaches to personalized federated learning, including meta-learning, local fine-tuning, and model interpolation.
Advanced machine learning methods are decisive in both
cutting-edge research and industrial practice. The aim of the course is for
students to move beyond the fundamental techniques and deepen their
understanding of state-of-the-art methods, models, and learning paradigms, and
to approach the design, implementation, and operation of robust, scalable
solutions with a systems-level mindset.
The course presents the advanced paradigms and the
systems-level way of thinking that are essential for designing and deploying
robust and scalable machine learning solutions. Particular attention is paid to
understanding the theoretical foundations, mastering the implementation of
advanced models, and critically assessing their performance and limitations.
Learning outcomes
Ez a tantárgy a KKK rendeletben meghatározott, következő kompetenciák fejlesztését szolgálja:
Knowledge
No learning outcomes recorded.
Skills
No learning outcomes recorded.
Attitudes
No learning outcomes recorded.
Autonomy and responsibility
No learning outcomes recorded.
Oktatási módszertan
Lectures and practical sessions, complemented by the study of the associated assigned materials (e.g., IPython notebooks, source code, descriptions, videos). The lectures and practical sessions may be held in person or online, in a synchronous or asynchronous format.
Tanulástámogató anyagok
Online források
Russell, S., & Norvig, P. (2021). Artificial Intelligence: a modern approach, 4th US ed. URL: https://aima.cs.berkeley.edu ; ; Chollet, F., Deep Learning with Python, Third Edition, Manning Publications, 2025; https://www.manning.com/books/deep-learning-with-python-third-edition; Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT press.; https://www.deeplearningbook.org/
Recommended preliminary knowledge for completing the subject
Knowledge type competencies
(azon előzetes ismeretek összessége, amelyek megléte nem kötelező, de a tantárgy eredményes teljesítését nagyban elősegíti)
nincs
Skill type competencies
(azon előzetes képességek és készségek összessége, amelyek megléte nem kötelező, de a tantárgy eredményes teljesítését nagyban elősegíti)
nincs
Recommended (non-compulsory) preliminary competencies
(azon ajánlott (nem kötelező) előzetesen megszerzendő kompetenciák összessége, amelyek jelentősen hozzájárulnak a tantárgy eredményes teljesítéséhez)
nincs
General rules
Requirements:
During the study period, students carry out homework/project work. Over the course of the semester, students must report on their progress. The completed solution, together with its documentation, must be submitted at the end of the semester.
A student may receive the end-of-semester signature recognising completion of the semester only if the submitted homework/project work meets the minimum requirements.
The course concludes with an examination. The end-of-semester grade is determined on the basis of the result achieved in the examination. A prerequisite for passing the examination (a grade of at least 'pass') is meeting the specified minimum requirement, namely achieving at least 40% of the maximum attainable score. Outstanding student performance during the study period may earn bonus points that count towards the examination result.
Additional possibilities:
The homework/project work may be made up until the end of the make-up (retake) week.
Assessment methods
In-term assessments
No detailed assessments provided.
Weight of in-term assessments
No weights provided.
Exam-period assessments
No detailed assessments provided.
Weight of exam elements
No weights provided.
Grade calculation
No grade thresholds provided.
Attendance requirements
No attendance requirements provided.
Rules for retake and resubmission
Not provided.
Short description
Not provided.
Detailed description
Not provided.
Recommended courses
Not provided.
Workload to complete the subject
No workload breakdown provided.
Validity of subject requirements
Requirements valid from:
—
Requirements valid until:
—
Curriculum placement
No curriculum placements recorded for this subject version.