- 🎓Education:
- MSc in Data Science @ University of Milano-Bicocca (2025 – present)
- BSc in Statistics and Information Management @ University of Milano-Bicocca (2022 – 2025)
- 📌Focus: Machine Learning, Statistical Modelling, Data Visualization, NLP, Deep Learning
- 🎯Particularly interested in: data analysis at the level of institutions and organizations (EU/Eurostat-style data, public opinion research)
- 🌍Location: Milan, 🇮🇹
| Area | Tools |
|---|---|
| Languages & Query | Python · R · SQL · Neo4j |
| Reports & Documentation | LaTeX · Overleaf |
| ML & DL | Scikit-learn · TensorFlow/Keras · Predictive Models · Neural Networks · Transfer Learning & Fine-tuning · XAI · Clustering |
| Statistical Modelling | GLM · Logistic Regression · Bootstrap Methods · Hypothesis Testing · PCA · and other methods |
| NLP | Text Preprocessing · Word Embeddings · Transformers (BERT/GPT-style) · Hugging Face · Sentiment Analysis · Topic Modelling · Text Classification |
| Ops & Workflow | Git/GitHub · VS Code · RStudio |
Also exposed to: SPSS, KNIME, MongoDB, SAS
| Project | What it does | Stack |
|---|---|---|
| Public-Perception-of-AI | Bachelor's thesis: sentiment and emotion analysis, topic modeling and SHAP on ~58K Reddit comments about AI | |
| Food-Image-Classification | Transfer learning vs. CNN from scratch on a small, imbalanced food dataset, with Grad-CAM, LIME and a live demo | |
| Warhammer-40K-Tournament-Analysis | Does army cost buy victory? API acquisition, SQLite database and statistical analysis of a world championship | |
| Mushroom-Edibility-ML-XAI | Hybrid KNIME-Python pipeline for mushroom classification with zero false negatives and SHAP explainability | |
| Europe-Aging-Population | Data visualization of Europe's demographic aging with Eurostat data | |
| Youth-Deprivation-Poset-Analysis | Poset-based identification of youth deprivation in 24 European countries, compared with a composite index | |
| Data-Mining-Portfolio-R | Four R projects on EDA, clustering, PCA and predictive modeling | |
| Trending-YouTube-Analysis | Fifteen pandas exercises on the Trending YouTube dataset across 10 countries |