Research
The Conecta 2030 Project leverages 5G, C-V2X, and AI to enhance road safety for vulnerable users, bringing together partners like TIM, Stellantis, USP, UFSCar, and THI in innovative, connected solutions.
The ML4LKD project leveraged machine learning and natural language processing to uncover latent knowledge in medical articles, producing impactful academic contributions.
Discover P2C, a classification technique that applies clustering to enable linear models to handle non-linear data effectively.
Discover ML-MDLText, a lightweight multilabel text classifier based on the minimum description length principle, offering incremental learning and label dependency handling.
The Gaussian Mixture Descriptors Learner (GMDL) is a lightweight multiclass online classifier, excelling in handling continuous features and large, dynamic datasets.
A dataset of 1,956 YouTube comments, labeled as spam or legitimate, supports research on spam filtering and machine learning in user-generated content.
Introducing MDLText, an advanced text classification tool leveraging the Minimum Description Length principle to deliver robust, efficient, and scalable performance for diverse applications.