Identifying Data 2023/24
Subject (*) Machine Learning II  Code 614544014
Study programme
Máster Universitario en Intelixencia Artificial
Descriptors Cycle Period Year Type Credits
Official Master's Degree 2nd four-month period
First Optional 3
Language
English
Teaching method Hybrid
Prerequisites
Department Ciencias da Computación e Tecnoloxías da Información
Coordinador
Fernández Blanco, Enrique
E-mail
enrique.fernandez@udc.es
Lecturers
Fernández Blanco, Enrique
E-mail
enrique.fernandez@udc.es
Web
General description A disciplina introduce ao alumnado nas técnicas de aprendizaxe automático aplicables en entornos que presentan restricións na distribución dos datos utilizados na xeración dos modelos: tratamento de fluxos, incorporación de novas experiencias, evolución dos conceptos ao longo do tempo ou a preservación da privacidade da información. A súa consideración require dunha capacitación específica na aplicación de técnicas de aprendizaxe incremental, detección de obsolescencias e confidencialidade na manipulación de conxuntos de datos.

1. Adquirir os coñecementos sobre o funcionamento das principais técnicas de aprendizaxe incremental.
2. Aplicar técnicas de aprendizaxe incremental para a análise de datos en tempo real en entornos estacionarios e non estacionarios.
3. Coñecer o principio de funcionamento dos principais paradigmas de aprendizaxe con preservación da privacidade.

Study programme competencies
Code Study programme competences
A11 CE10 - Ability to implement, validate and apply a stochastic model starting from the observed data on a real system, and to perform a critical analysis of the obtained results, selecting those ones most suitable for problem solving
A12 CE11 - Understanding and command of the main techniques and tools for data analysis, both from the statistical and the machine learning viewpoints, including those devised for large volumes of data, and ability to select those ones most suitable for problem solving
A13 CE12 - Ability to outline, formulate and solve all the stages of a data project, including the understanding and command of basic concepts and techniques for information search and filtering in big collections of data
A16 CE15 - Knowledge of computer tools in the field of machine learning and ability to select those ones most suitable for problem solving
B2 CG02 - Successfully addressing each and every stage of an AI project
B3 CG03 - Searching and selecting that useful information required to solve complex problems, with a confident handling of bibliographical sources in the field
B4 CG04 - Suitably elaborating written essays or motivated arguments, including some point of originality, writing plans, work projects, scientific papers and formulating reasonable hypotheses in the field
B5 CG05 - Working in teams, especially of multidisciplinary nature, and being skilled in the management of time, people and decision making
B6 CB01 - Acquiring and understanding knowledge that provides a basis or opportunity to be original in the development and/or application of ideas, frequently in a research context
B7 CB02 - The students will be able to apply the acquired knowledge and to use their capacity of solving problems in new or poorly explored environments inside wider (or multidisciplinary) contexts related to their field of study
B8 CB03 - The students will be able to integrate different pieces of knowledge, to face the complexity of formulating opinions (from information that may be incomplete or limited) and to include considerations about social and ethical responsibilities linked to the application of their knowledge and opinions
B9 CB04 - The students will be able to communicate their conclusions, their premises and their ultimate justifications, both to specialised and non-specialised audiences, using a clear style language, free from ambiguities
C3 CT03 - Use of the basic tools of Information and Communications Technology (ICT) required for the student's professional practice and learning along her life
C4 CT04 - Acquiring a personal development for practicing a citizenship under observation of the democratic culture, the human rights and the gender perspective
C7 CT07 - Developing the ability to work in interdisciplinary or cross-disciplinary teams to provide proposal that contribute to a sustainable environmental, economic, political and social development
C8 CT08 - Appreciating the importance of research, innovation and technological development in the socioeconomic and cultural progress of society
C9 CT09 - Being able to manage time and resources: outlining plans, prioritising activities, identifying criticisms, fixing deadlines and sticking to them

Learning aims
Learning outcomes Study programme competences
To acquire knowledge of how the main incremental learning techniques work. AC10
AC11
AC12
AC15
BC2
BC3
BC4
BC5
BC6
BC7
BC8
BC9
CC3
CC4
CC7
CC8
CC9
To apply incremental learning techniques for the analysis of real-time data in stationary and non-stationary environments AC10
AC11
AC12
AC15
BC2
BC3
BC4
BC5
BC6
BC7
BC8
BC9
CC3
CC4
CC7
CC8
CC9
To know the working principle of the main privacy-preserving learning paradigms AC10
AC11
AC12
AC15
BC2
BC3
BC4
BC5
BC6
BC7
BC8
BC9
CC3
CC4
CC7
CC8
CC9

Contents
Topic Sub-topic
1. Theory 1. Machine Learning Online
2. Concept Drift
3. Federated Learning
2. Practice 1. Machine Learning Online and Concept Drift
2. Federated Learning

Planning
Methodologies / tests Competencies Ordinary class hours Student’s personal work hours Total hours
Guest lecture / keynote speech A11 A12 A13 A16 B3 B6 B7 B8 B9 C3 C4 C8 10 10 20
Seminar A11 A12 A13 A16 B2 B4 B5 C7 C9 4 20 24
ICT practicals A11 A12 A13 A16 B3 B6 B7 B8 C3 C4 C8 7 21 28
Mixed objective/subjective test A11 A12 A13 A16 B4 B6 B7 1 0 1
 
Personalized attention 2 0 2
 
(*)The information in the planning table is for guidance only and does not take into account the heterogeneity of the students.

Methodologies
Methodologies Description
Guest lecture / keynote speech The contents of the course will be taught indistinctly between lectures and interactive classes. The completion of all the proposed activities is necessary, as well as the attendance to all the classes (lectures and interactive) to pass the course.

Expository classes (theory): will consist of the explanation of the different sections of the course syllabus, with the help of electronic media (presentations, videos, etc.)
Seminar Case studies: students may be presented with real or fictional work scenarios that present certain problems. Students will have to apply the theoretical and practical knowledge of the subject to find a solution to the question or questions posed. As a general rule, case studies will be carried out in groups. The different working groups will present and share their solutions.
ICT practicals Interactive classes (practical): different practical problems related to the content of the subject will be posed for the student to solve individually or in groups.
Project-based learning: students may be given practical projects whose scope requires them to dedicate a significant part of their time to the subject.
Autonomous work: the scope and objectives of the projects, use cases and/or practical problems may require autonomous work on the part of the students, albeit under the supervision of the teaching staff.
Mixed objective/subjective test A mixed test which can contain quiz questions, short=answer questions or development questions. It is going to evaluate the teorethical part of the subject and it can contain questions about the content of the seminars or practical exercises

Personalized attention
Methodologies
Guest lecture / keynote speech
ICT practicals
Seminar
Description
Office hours: Office hours will be used to solve students' doubts related to the contents of the subject. These office hours can be both face-to-face and virtual (via email, virtual campus or Microsoft Teams platform).

Virtual Classroom: This subject will have a virtual classroom where students will be provided with all the necessary material in digital format. Different communication tools will also be provided to support both teaching and office hours, including videoconferencing, chat, e-mail, forums...

Assessment
Methodologies Competencies Description Qualification
Mixed objective/subjective test A11 A12 A13 A16 B4 B6 B7 Subjective test which can be a mixture model with quiz questions and some short-answer or development questions. 50
ICT practicals A11 A12 A13 A16 B3 B6 B7 B8 C3 C4 C8 This mark includes the evaluation exercises made during the practical lessons and the developed project. 30
Seminar A11 A12 A13 A16 B2 B4 B5 C7 C9 This is going to include the grading og the practical exercises and the proyects developed in the seminars. 20
 
Assessment comments
In order to pass the course, the student will have to carry out all the proposed activities and pass the corresponding exams.

First opportunity:
To pass the subject, the student must deliver and pass the proposed activities (50% of the final grade) and pass the final exam (50% of the grade).

Mid-term exams:
No mid-term exams will be held.

Second opportunity:
The grade obtained in the laboratory practices during the course is maintained, as well as its weight in the final grade. Students who have notreached the cut-off mark in the activities proposed during the previous call, may submit, prior to the second chance final exam, similar activities,which will be proposed by the teachers. Once both parts have been passed separately, the exam will account for the 50% of the final mark and thelaboratory practices for the remaining 50%.

Exemption from attendance:
In case of dispensation of attendance, students will be examined under the same conditions as students in the first round. 

Repeating students:
In case of repeating students, they will be examined under the same conditions as students in the first round.

No-show qualification:
The student will receive the qualification of "no-show" when he/she does not take the final exam.

Fraudulent performance of exercises or tests: 
The fraudulent execution of tests or assessment activities, once proven, will result in a direct failing grade in the examination in which it was committed. The student will be given a grade of "suspenso" (numeric grade 0) in the corresponding academic year's examination, whether the offense occurs in the first opportunity or the second. In order to do so, the student's grade will be modified in the first opportunity's record, if necessary.

Evaluation of competences: 
In general, the development of the practical activities, projects and use cases, as well as the preparation of the theoretical topics will allow students to work on the basic, general and transversal competences of the subject. Specifically, through the projects and use cases, the competences CT7, CT9, CG5, CG4, CG2 will be assessed. The development of the practices, as well as the final test, will allow the evaluation of the specific competences: CE10, CE11, CE12, CE15.

Equality:

According to the various applicable regulations for university teaching, the gender perspective should be incorporated into this subject (using non-sexist language, using bibliographic references from authors of both genders, encouraging the participation of male and female students in class).

Efforts will be made to identify and modify prejudices and sexist attitudes, and the environment will be influenced to change them and promote values of respect and equality.

Situations of gender-based discrimination should be identified, and actions and measures should be proposed to correct them.


Sources of information
Basic Bahri, M., Bifet, A., Gama, J., Gomes, H. M., & Maniu, S (2021). Data stream analysis: Foundations, major tasks and tools. Wiley nterdisciplinary Reviews: Data Mining and Knowledge Discovery,11(3)
Orabona, F. (2019). A modern introduction to online learning.. arXivpreprint arXiv:1912.13213
Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M., & Bouchachia, A. (2014). A survey on concept drift adaptation.. CM computing surveys(CSUR),46(4), 1-37
Li, T., Sahu, A. K., Talwalkar, A., & Smith, V. (2020). Federated learning: Challenges, methods, and future directions.. IEEE signal processing magazine, 37(3), 50-60
Yang, Q., Liu, Y., Chen, T., & Tong, Y. (2019). Federated machine learning: Concept and applications.. ACM Transactions on Intelligent Systems and Technology (TIST), 10(2), 1-19
Lu, J., Liu, A., Dong, F., Gu, F., Gama, J., & Zhang, G. (2018). Learning under concept drift: A review.. IEEE Transactions on Knowledge and DataEngineering,31(12), 2346-2363
Bifet, A., Gavalda, R., Holmes, G., & Pfahringer, B (2018). Machine learning for data streams: with practical examples in MOA. MIT Press
Gomes, H. M., Read, J., Bifet, A., Barddal, J. P., & Gama, J. (2019). Machine learning for streaming data: state of the art, challenges, and opportunities.. ACM SIGKDD Explorations Newsletter,21(2), 6-22
Hoi, S. C., Sahoo, D., Lu, J., & Zhao, P. (2021). Online learning: A comprehensive survey. Neurocomputing,459, 249-289.

Complementary Gama, J., Medas, P., Castillo, G., & Rodrigues, P. (2004). Learning with drift detection. InBrazilian symposium on artificialintelligence(pp. 286-295). Springer, Berlin, Heidelberg.
AbdulRahman, S., Tout, H., Ould-Slimane, H., Mourad, A., Talhi, C., & Guizani, M. (2020). A survey on federated learning: The journey fromcentralized to distributed on-site learning and beyond.. IEEE Internet of Things Journal, 8(7), 5476-5497
Bifet, A., & Gavalda, R. (2009). Adaptive learning from evolving data streams.. InAdvances in Intelligent Data Analysis VIII
McMahan, B., Moore, E., Ramage, D., Hampson, S., & y Arcas, B. A. (2017). Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics (pp. 1273-1282).
Rahman, K. J., Ahmed, F., Akhter, N., Hasan, M., Amin, R., Aziz, K. E., ... & Islam, A. N. (2021). hallenges, applications and design aspects of federated learning: A survey.. IEEE Access,9, 124682-124700.
Bifet, A., Gavalda, R. (2007). Learning from time-changing data with adaptive windowing. Proceedings of the 2007 SIAM international conference ondata mining, pp. 443-448. Society for Indust
Gama, J., & Castillo, G. (2006). Learning with local drift detection.. Advanced Data Mining and Applications: Second International Conference,ADMA 2006, Xi’an, China, Augu
Gomes, H. M., Montiel, J., Mastelini, S. M., Pfahringer, B., & Bifet, A. (2020). On ensemble techniques for data stream regression. In 2020International Joint Conference on Neural Networks (IJCNN) (pp. 1-8)
Ghesmoune, M., Lebbah, M., & Azzag, H (2016). State-of-the-art on clustering data streams.. Big Data Analytics, 1, 1-27
(). ttps://federated.withgoogle.com/.


Recommendations
Subjects that it is recommended to have taken before
Machine Learning I  /614544012

Subjects that are recommended to be taken simultaneously

Subjects that continue the syllabus

Other comments

The students should be familiar with mid-level programming concepts, linear algebra, calculus and stadistics. The knowledge of basic concurrence and parallel architecture is also helpful

Equality:

According to the various applicable regulations for university teaching, the gender perspective should be incorporated into this subject (using non-sexist language, using bibliographic references from authors of both genders, encouraging the participation of male and female students in class).

Efforts will be made to identify and modify prejudices and sexist attitudes, and the environment will be influenced to change them and promote values of respect and equality.

Situations of gender-based discrimination should be identified, and actions and measures should be proposed to correct them.



(*)The teaching guide is the document in which the URV publishes the information about all its courses. It is a public document and cannot be modified. Only in exceptional cases can it be revised by the competent agent or duly revised so that it is in line with current legislation.