statistics

Meetup Nubank: I talked about my journey as a data scientist in different sectors and how I used various techniques, from linear regressions to more complex methods like neural networks, in a creative and practical way.

We built a data-driven framework for cable fault detection. We validated our approach using 60 random labeled files, achieving 98% for both accuracy and F1-score.

PyData Global: A 2‑hour workshop on machine learning pipelines at PyData Global, focusing on data leakage prevention and efficient model building with scikit-learn and other tools.

PyData SP: Workshop highlighting data science techniques for analyzing categorical data, such as chi‑square, Cramér’s V, CA, MCA, entropy, information gain, and so on.