| Code |
16791
|
| Year |
2
|
| Semester |
S2
|
| ECTS Credits |
6
|
| Workload |
PL(30H)/T(30H)
|
| Scientific area |
ENGENHARIA INFORMÁTICA
|
|
Entry requirements |
N/A
|
|
Learning outcomes |
At the end of the course, students should be able to:
* Understand concepts associated with Data Science, namely the distinction between data, information and knowledge, different storage formats, and the application of data cleaning and normalisation techniques;
* Analyse and interpret data using Data Science methods, namely collection and integration techniques, exploratory analysis and statistical and visualisation methods, with a view to knowledge extraction;
* Know and apply the main IA learning paradigms, namely supervised, unsupervised and reinforcement learning, understanding their theoretical foundations and scope of application;
* Know and explore artificial intelligence techniques in the areas of search, generative AI and computer vision, understanding the theoretical foundations and potential application of each approach.
|
|
Syllabus |
1. Fundamentos de Dados: dados e tipos, informação vs conhecimento, viés na recolha, formatos CSV/JSON/XML/Parquet.
2. Recolha e Ingestão: APIs, bases de dados, web scraping, ETL/ELT, integração, data cleaning (missing values, outliers, inconsistências, escalas).
3. Análise e Preparação: análise exploratória, estatísticas, visualização, normalização (máx., Min-Max, Z-score, IQR).
4. Análise Preditiva: introdução, métodos, métricas.
5. Algoritmos de Pesquisa: DFS, BFS, gulosa, A*, MinMax, Alpha-Beta, profundidade.
6. Aprendizagem Supervisionada: função de custo, gradiente, redes neuronais.
7. Aprendizagem Automática: generalização, protocolos, validação cruzada.
8. Aprendizagem Não Supervisionada: clustering, K-means, agglomerative.
9. Aprendizagem por Reforço: Q-Learning.
10. IA Generativa: BoW, TF-IDF, embeddings, transformers, LLMs.
11. Visão Computacional: classificação, CNNs, VLMs, foundational models.
|
|
Main Bibliography |
[1] D.P. Kroese, Z.I. Botev, T. Taimre, R. Vaisman. "Data Science and Machine Learning: Mathematical and Statistical Methods." Chapman & Hall/CRC, 2019.
[2] Hui Lin, Ming Li, Practitioner’s Guide to Data Science, 2023.
[3] Joel Grus. Data Science from Scratch, 2nd Edition. O'Reilly Media, Inc., 2019.
[4] Andriy Burkov. The Hundred-Page Machine Learning Book, 2019.
[5] S. Russell and P. Norvig; Artificial Intelligence: A Modern Approach, Pearson, 2021.
[6] Richard E. Korf. Artificial intelligence search algorithms. Algorithms and theory of computation handbook: special topics and techniques (2nd ed.). Chapman & Hall/CRC, 2010.
|
|
Teaching Methodologies and Assessment Criteria |
The contents of this course are delivered in theoretical classes (expository and interactive method) and their practical component is explored in laboratory practical classes, through the completion of laboratory guides. The theoretical and practical components are assessed through a Theoretical Test (T), composed of direct and essay-style theoretical questions, as well as practical exercises to solve. Thus, the final grade (FG) is: FG = T.
A student passes if they obtain a grade equal to or greater than 9.5. If a student passes during the teaching-learning period, they are exempt from the exam, although they may choose to take it. The exam replaces the Theoretical Test (T) grade in case of improvement. In sum:
* FG < 5.5 (out of 20) => Failed and Not Admitted to Exam;
* FG >= 9.5 (out of 20) => Passed and Exempt from Exam;
* Remaining cases => Failed and Admitted to Exam.
|
|
Language |
Portuguese. Tutorial support is available in English.
|