In order to pass the subject, the student must obtain a minimum score of 5 out of 10 in the result of combining the grades of the objective test, the laboratory practices and the supervised works. In addition, the student must obtain a minimum score of 2 out of 5 points in the objective test. If the student does not obtain this minimum grade, the grade of the subject will be that corresponding to the grade of the objective test.

In the second opportunity, the grade obtained in the laboratory practices and supervised works will be maintained, not being able to obtain again a grade since it results from the continuous evaluation of the work during the credits of practice of the subject. The student can retake the examination of the objective test, the criteria for obtaining the total score being those indicated at the beginning of this section.

Part-time students must deliver in their reports on the same dates as full-time students, and attend the RGTs in which they will be corrected. Similarly, it is recommended that they attend the practice classes.

Learning outcomes	Study programme competences
Understand the relationship between the complexity of learning models, training data features and overfitting, and know the mechanisms to avoid it.	A24 A25
Develop skills to design the stages of a complete data analysis process based on automatic learning techniques.		B2 B7 B9 B10	C1
Know how to correctly apply automatic learning techniques to obtain reliable and significant results.	A24	B3 B8
Know the most representative and current techniques of unsupervised, semi-supervised and supervised learning, with and without reinforcement.	A24	B8
Know the most representative learning techniques for the classic problems of classification, regression and clustering, and other less classic ones such as sorting problems, one class problems or multitasking.	A24	B8
Identify appropriate data analysis techniques according to the problem.	A25	B3 B8
Manage the most current tools and work environments in the field of machine learning.	A26	B2 B10

Topic	Sub-topic
1. Introduction	1.1. Introduction to Machine Learning 1.2. Inductive Learning
2. Supervised learning	2.1. Introduction 2.2. Artificial Neural Networks 2.3. Support Vector Machines 2.4. Decision trees 2.5. Regression trees and regression model trees 2.6. Instance-based learning
3. Evolutionary Computation	3.1. Genetic Algorithms 3.2. Genetic Programming 3.3. Swarms and other Evolutionary Computation techniques
4. Methodologies in data analysis	4.1. Training, evaluation and model selection methodologies 4.2. Methodologies of a data analysis project
5. Unsupervised learning	5.1. Clustering methods 5.2. Self-organised networks

Methodologies / tests	Competencies	Ordinary class hours	Student’s personal work hours	Total hours
Guest lecture / keynote speech	A24 A25 B3 B8 B9	30	60	90
Laboratory practice	A26 B2 B3 B10 C1	18	18	36
Supervised projects	B2 B3 B7 B9 B10	10	10	20
Objective test	A24 A25 B8 B9	2	0	2

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	Description
Guest lecture / keynote speech	Theoretical teaching of the subject matter of the course
Laboratory practice	Solve practical problems by using the different techniques that will be explained in the theory classes.
Supervised projects	Writing, under the supervision of the teacher, of the reports explaining the resolution of the problems carried out in the laboratory practices and the results obtained.
Objective test	This is a written assessment test in which the student must demonstrate the knowledge acquired from the subject.

Methodologies	Competencies	Description	Qualification
Laboratory practice	A26 B2 B3 B10 C1	Resolution of real world problems using the methodology, for which several techniques explained in theory will be used, and the student will be stimulated to generate new ideas for the resolution of these problems.	25
Supervised projects	B2 B3 B7 B9 B10	Writing of the report on the resolution of the real problems carried out in the laboratory practices. The writing of the report will include a bibliographic review of the most important works related, written in English for the most part, documentation on the problem to be solved, methodology used, and comparison of the results found in the application of the different techniques, as well as a critical evaluation of both the results obtained and the information used.	25
Objective test	A24 A25 B8 B9	Test questions about the contents of the course, based on the different machine learning techniques and their applications.	50

Assessment comments
In order to pass the subject, the student must obtain a minimum score of 5 out of 10 in the result of combining the grades of the objective test, the laboratory practices and the supervised works. In addition, the student must obtain a minimum score of 2 out of 5 points in the objective test. If the student does not obtain this minimum grade, the grade of the subject will be that corresponding to the grade of the objective test. In the second opportunity, the grade obtained in the laboratory practices and supervised works will be maintained, not being able to obtain again a grade since it results from the continuous evaluation of the work during the credits of practice of the subject. The student can retake the examination of the objective test, the criteria for obtaining the total score being those indicated at the beginning of this section. Part-time students must deliver in their reports on the same dates as full-time students, and attend the RGTs in which they will be corrected. Similarly, it is recommended that they attend the practice classes.

Basic	D. Borrajo, J. González, P. Isasi (2006). Aprendizaje automático. Sanz y Torres Basilio Sierra Araujo (2006). Aprendizaje automático: conceptos básicos y avanzados. Aspectos prácticos utilizando el software WEKA. Pearson Education Marcos Gestal, Daniel Rivero, Juan Ramón Rabuñal, Julián Dorado, Alejandro Pazos (2010). Introducción a los Algoritmos Genéticos y a la Programación Genética. Servicio de Publicaciones de la Universidade da Coruña Ethem Alpaydin (2004). Introduction to Machine Learning. MIT Press David Aha (). Lazy Learning. Kluwer Academics Publishers T.M. Mitchell (1997). Machine Learning. McGraw Hill Saso Dzeroski, Nada Lavrac (). Relational Data Mining. Springer Andrew Webb (2002). Statistical Pattern Recognition. Wiley

Complementary

Other comments