Bruno Vaz is a Data Scientist/Engineer working in a Data & AI consulting company. He received a bachelor’s degree in Mathematics, with a minor in Informatics, and is finishing his master’s degree in Data Science from the University of Porto. His research interests include generative deep learning, with a special interest in generative adversarial networks, and synthetic data generation.
Bruno’s participation in the Future Talent Programme (FTP22) was supported by FCCN Portugal.
Lightning Talk Topic
Fake News Detection Models — Introducing GAN Generated Synthetic Samples to Improve Performance
News datasets are extremely imbalanced, with the fake news not being as well represented as the real news. Thus, machine learning models for fake news detection do not perform as good as expected. Hence, Generative Adversarial Networks can be used to produce high-quality synthetic samples to better represent the fake news data, improving the models’ performances.