Prof. Rafael S. de Souza
Research Profile
📧 [email protected]
Affiliation: University of Hertfordshire
Lecturing at: Instituto de Astronomia, Geofísica e Ciências Atmosféricas (IAG), Universidade de São Paulo – 2025
This course offers a rigorous yet accessible introduction to the use of statistical and machine learning techniques in astronomy. From foundational concepts to advanced data-driven methodologies, students will explore how modern astronomical research increasingly relies on tools from statistics, artificial intelligence, and data science.
Through a blend of theoretical derivations and practical coding exercises, participants will gain firsthand experience with methods that underpin contemporary astrophysical discovery. While examples are primarily drawn from astronomy, the course content is broadly applicable to students in Physics and related fields who seek to deepen their understanding of data-driven inference.
- Generalized Linear Models (GLMs)
- Random Forest, Generalized Additive Models
- Neural Networks (MLPs, CNNs, RNNs)
- Dimensionality Reduction
- Clustering & Unsupervised Learning
- GMMs
- QRPCA, de Souza 2022
- Deep Learning with Keras (2018)
- Goodfellow et al., Deep Learning (2016)
- Bishop, Pattern Recognition and Machine Learning (2006)
- Srivastava et al., "Dropout" (2014)
Click the Binder badge above ☝️ to launch the course environment in your browser — no installation required.
To run locally:
git clone https://github.com/RafaelSdeSouza/astrostats-2025.git
cd astrostats-2025
pip install -r requirements.txt
jupyter notebook