Data Science MSc student at Tilburg University · Bayesian modeling · Machine learning · Data-driven policy
I use statistics and machine learning to answer questions that matter for decision-makers, from public policy to business strategy. My background combines management and marketing (Bocconi University) with data science (Tilburg University), and I have studied and worked in Italy, the Netherlands, India and China.
MSc thesis: a machine learning early-warning model for severe youth population decline in rural Spanish municipalities, designed to inform EU regional and demographic policy and evaluated with province-level cross-validation.
| Project | What I did | Stack |
|---|---|---|
| GHG emissions: Bayesian multilevel model | Validated and aggregated corporate emissions data across sectors and countries, built multilevel Bayesian regressions and compared specifications via cross-validation | R · brms |
| Property value prediction challenge | Cleaned and engineered features from property transaction data, built an ensemble of XGBoost, LightGBM and CatBoost with blend weights selected via 5-fold cross-validation | Python · scikit-learn |
- Modeling: Bayesian statistics (brms), machine learning (scikit-learn, LightGBM, XGBoost, CatBoost)
- Data: Python (pandas), R (dplyr, ggplot2), SQL
- Visualization and analytics: Tableau, Google Analytics 4, Excel
- Research methods: Social Network Analysis, survey design, cross-validation
- Workflow: Git and GitHub
- Product Marketing intern at Sky Italia: commercial performance analysis with Tableau and Excel dashboards
- MSc in Marketing Management and BSc in Business Administration at Bocconi University, exchange semester at IIM Bangalore
Open to data science and analytics opportunities in the Netherlands. LinkedIn · fabiopugno21@gmail.com