← Back to curriculum
Module 3
Workload
105h
Disciplines
4
Full program
360h
Predictive Models
Core module focused on building predictive models and modern AI applications.
The third module concentrates the predictive core of the program. It combines statistical and computational fundamentals with machine learning, deep learning, and natural language processing techniques applied to real problems.
Module disciplines
01
30h
Supervised Learning
Professor Prof. Dr. Carlos Ferreira
Syllabus
Learn more →
Linear models for regression and classification. Logistic regression. Evaluation of predictive models. Naive Bayes model. Tree-based models. Additive models (ensembles). Support Vector Machines. Kernel trick.
02
30h
Artificial Neural Networks and Deep Learning
Professor Sarah
Syllabus
Learn more →
Concepts and modeling of artificial neurons and artificial neural networks. Perceptron. Multilayer Perceptron: architecture, activation functions, cost functions, training with the backpropagation algorithm, and applications to regression and classification tasks. Deep neural networks: foundations, architecture examples, and modern applications for text and image processing.
03
15h
Ensemble Methods
Professor Prof. Dr. Matheus Haddad
Syllabus
Learn more →
Ensemble tasks in machine learning. Types of ensemble methods. Ensemble methods for classification. Hyperparameter tuning techniques for optimizing ensemble methods.
04
30h
Natural Language Processing (Transformers + LLMs)
Professor Prof. Dr. Filipe Nunes Ribeiro
Syllabus
Learn more →
Introduction to Natural Language Processing (NLP). Text preprocessing techniques. Foundations of semantic and vector text representation techniques. Notions of deep learning architectures for NLP: sequential models, Transformers, and large language models (LLMs).