Unsupervised machine learning

The objective of this course is to equip students with the ability to identify and appropriately select the most suitable unsupervised learning techniques to solve specific problems within various domains. To achieve this, students must understand the distinctive characteristics of various unsupervised learning techniques, know how to evaluate their strengths and limitations, and determine which technique is most appropriate based on the type of data, its distribution, and the context of the problem to be solved.

- Topic 1: Introduction to Unsupervised Learning
- Topic 2: Evaluation and Validation in Unsupervised Learning
- Topic 3: Dimensionality Reduction
- Topic 4: Unsupervised Methods Based on Neural Networks
- Topic 5: Clustering Techniques
- Topic 6: Association Techniques and Pattern Analysis

Basic Bibliography:
- Notes on Unsupervised Machine Learning. Handouts provided by the faculty covering all course topics.
- Hastie, Trevor, Jerome Friedman, and Robert Tibshirani. The Elements of Statistical Learning Data Mining, Inference, and Prediction. New York, NY: Springer New York, 2001. Available online.
- Bishop, Christopher M. Pattern Recognition and Machine Learning. New York: Springer, 2006. Print. Available at ETSE library (C50 721).

Supplementary Bibliography:
- Larose, Daniel T, and Chantal D Larose. Data Mining and Predictive Analytics. Second edition. Hoboken, New Jersey: John Wiley & Sons, 2015. Available online.
- Johnston, Benjamin, Aaron Jones, and Christopher Kruger. Applied Unsupervised Learning with Python: Discover Hidden Patterns and Relationships in Unstructured Data with Python. 1st edition. Birmingham, UK: Packt Publishing, 2019. Available online.

The course contributes to the following competencies:

Basic:
- [CB3] The ability to gather and interpret relevant data (usually within their field of study) to make judgments that include reflection on relevant social, scientific, or ethical issues.

General:
- [CG3] Ability to design and create high-quality models and solutions based on Artificial Intelligence that are efficient, robust, transparent, and responsible.
- [CG4] Ability to select and justify the appropriate methods and techniques to solve a specific problem or to develop and propose new methods based on artificial intelligence.
- [CG5] Ability to conceive new computational systems and/or evaluate the performance of existing systems that integrate artificial intelligence models and techniques.

Specific:
- [CE1] Ability to use mathematical and statistical concepts and methods to model and solve artificial intelligence problems.
- [CE12] Understand the foundations of algorithms and models in artificial intelligence to solve complex problems, understand their computational complexity, and have the ability to design new models.
- [CE15] Know and be able to apply and explain correctly the validation techniques for artificial intelligence solutions.

Transversal:
- [TR1] Ability to communicate and transmit their knowledge, skills, and abilities.
- [TR2] Ability to work in a team, in interdisciplinary environments, and manage conflicts.

Learning Outcomes
- Ability to select different unsupervised learning techniques to solve a problem in a specific domain.
- Knowledge and application of clustering techniques based on similarity criteria between data.
- Understanding and application of dimensionality reduction techniques through feature selection and extraction.

The teaching methodology is focused on the theoretical and practical aspects of unsupervised learning and how it differs from other types of learning. Considering this, two types of learning activities are distinguished: theoretical classes and practical classes.

- Theoretical Classes: 25 hours of lectures will be given in 1-hour sessions. These are aimed at explaining the concepts that support this learning paradigm.
- Practical Classes: 15 hours of practical classes will be conducted in the computer lab in 10 sessions of 1.5 hours each, allowing students to familiarize themselves practically with the issues discussed in the theoretical classes. Attendance at these classes is MANDATORY.

Face-to-Face Training Activities and Their Relation to the Degree Competencies
- Theoretical classes given by the professor and seminar presentations. Competencies worked on: CE1, CE12, CE15, CG3, CG4.
- Laboratory practical classes, problem-solving, and practical cases. Competencies worked on: CE1, CE12, CE15, CB3, CG3, CG4, CG5, TR1, TR2.
- Exam. Competencies worked on: CE1, CE12, CE15, CG3, CG4, CG5.

Non-Face-to-Face Training Activities and Their Relation to the Degree Competencies
- Personal work of the student: bibliography consultation, autonomous study, development of programmed activities, preparation of presentations and papers. Competencies worked on: CE1, CE12, CE15, CG3, CG4, CG5.

Ordinary Opportunity:
- Practical assignments: 40%
- Final exam: 60%

Students will address the resolution of various problems proposed in the computer lab. They must complete assignments presenting the obtained results. Several of these assignments will be mandatory and others optional. All assignments must be submitted before the specified dates and must meet minimum quality requirements to be considered. Acquired knowledge, compliance with specifications, methodology, rigor, and presentation of results will be evaluated.

Provided that the minimum attendance requirements are met, a total score of 5 or higher must be achieved in both practical assignments and the final exam to pass the course. It is essential to have submitted all mandatory assignments to pass.

Extraordinary Opportunity:
- Provided that the minimum attendance requirements are met, it will be possible to retake the parts that were not passed in the ordinary opportunity: unsubmitted assignments, practical assignments, and the exam.

Condition of Not Submitted:
- Students who have not been evaluated in any aspect of the course. Additionally, students with a percentage that does not represent more than 10% of the maximum total grade for the course can also opt for the Not Submitted condition, for which they must inform the course coordinator.

Repeating Students:
- Evaluations are not carried over between academic years.

Attendance Control
- Attendance at practical sessions is mandatory. If the student attends less than 80% of the practical sessions, they will be considered to have not passed the course.

In case of fraudulent completion of exercises or tests, the Academic Performance Evaluation and Qualification Review Regulations will be applied. According to the ETSE Plagiarism Regulations (approved by the ETSE Board on 19/12/2019), total or partial copying of any practical or theoretical exercise will result in failure in both course opportunities, with a grade of 0.0 in both cases.

Work in the Classroom:
- Theory classes: 25 hours
- Practical classes: 15 hours
- Individual tutoring: 1 hour

Total hours of classroom work: 40 hours

Personal work of students (study, completion of exercises, practical assignments, projects) and other activities (non-classroom evaluation): 71.5 hours

Due to the strong interrelationship between the theoretical and practical parts and the progression in the presentation of closely related concepts in the theoretical part, it is recommended to dedicate daily study or review time.

The USC virtual campus will be used for all teaching, publication of materials, practical scripts, and assignment submissions. The preferred language for delivering expository and interactive classes is Spanish, but some content in the bibliography and notes may be in English.