TDA-Brio-Wu_MHD_CNN-XXXIX_EFNNE
Pipeline using topological data analysis to detect and quantify the failure modes of neural temporal predictors of MHD simulations of the Brio-Wu shock tube problem.
Software
A selection of open-source codes and libraries I maintain or co-develop. Most are hosted on my GitHub profile.
Pipeline using topological data analysis to detect and quantify the failure modes of neural temporal predictors of MHD simulations of the Brio-Wu shock tube problem.
Pipeline using topological data analysis to study gravitational-wave-like signals at very low signal-to-noise ratio, developed for the 4th School on Data Science & Machine Learning.
Unsupervised topological and contrastive representation learning for Galaxy Zoo 2, combining persistence diagrams, self-supervised embeddings, and clustering.
Python Modules for a Static Star Model: a code and teaching resource for modeling the internal structure of a static, non-magnetic star, accompanied by theoretical documentation.
An AI-assisted workflow for literature reviews using retrieval-augmented generation, with integration to arXiv and a Streamlit-based interface.
Collection of Jupyter notebooks for PET–Física/UFRN covering topics such as astronomy, fluid dynamics, quantum mechanics, and dynamical systems, with a strong emphasis on numerical methods.
Material for the minicourse “Mathematical Foundations of Machine Learning”, including notebooks and exercises on theory of supervised/unsupervised learning.
Material for the minicourse “Uma breve introdução ao Machine Learning”, including notebooks and exercises on supervised/unsupervised learning, tensor computation, and dimensionality reduction.
Introductory Python course tailored for Chemical Engineering students, focusing on NumPy, SciPy, and common scientific computing workflows.