Machine Learning for Management Research
İstanbul Bilgi University — two-day hands-on workshop (Python) · July 5–6, 2026

This page hosts all materials for the two-day workshop on machine learning for management and social-science research. The workshop covers the foundations of predictive modeling — train/test evaluation and cross-validation, regularization and variable selection, tree-based methods and gradient boosting, and model interpretation with SHAP and partial dependence — applied to real published studies. Every hands-on tutorial is a Google Colab notebook that runs in the browser with a free Google account; all data load automatically from public links, so nothing needs to be installed or uploaded.

Day 1 — Foundations, regression & variable selection (PDF)

Day 2 — Learners & interpretation (SVM, trees/forests, XGBoost, SHAP) (PDF)

Day 1 — foundations, regression, regularization

	Voter turnout & term limits — OLS and out-of-sample prediction
	Economic shocks & regional elite splits — the Catalan Lliga (Vall-Prat 2022) — regularization + ABESS
	Election-law reform after 2000 (Palazzolo & Moscardelli) — regularization + ABESS
	Determinants of perceived ageism (Du et al. 2025) — reproduce + LASSO selection
	Advance-care-planning engagement (Han et al. 2025) — reproduce + regularization

Day 2 — tree-based learners & interpretation

	Militarized disputes — random forest / XGBoost + SHAP
	Pinyon Jay habitat — RF/XGBoost + SHAP + partial dependence
	Snake occurrence — RF/XGBoost + SHAP + partial dependence
	Civil-war onset (Fearon–Laitin) — SVM, k-NN & random forest + SHAP