Home » Departments and Centres » Academic catalogue » Course detail

Open sidebar menu

Text Analysis and Causal Machine Learning (ECO-AD-TEXTCAUS)

ECO-AD-TEXTCAUS

Description

Course overview:

Session 1:
ML recap, interpretability and Image models
•    Conditional Expectation function: Prediction vs. explanation in social science
•    Model evaluation (cross-validation, overfitting)
•    Regularized Regression models
•    CART models
•    Explainability tools (feature importance, SHAP)

Session 2:
Text as data – Part 1
•    Word embeddings (word2vec) 
•    Text pre-processing
•    The “Geometry of culture” approach
•    Applications

Session 3:
Text as data – Part 2
•    The Encoder revolution: from low memory to BERT
•    Transformer models: evolution towards GPT
•    Prompting strategies
•    Risks: hallucinations, bias, guard-rails. 

Session 4:
Causality Meets Machine Learning (Part 1)
•    Partialling Out and Neyman orthogonality
•    ML for causal inference (Double Post-Lasso, DDML)
•    Applications

Session 5:
Causal ML: heterogeneity, observed, unobserved confounders (Part 2)
•    DML-IV
•    Heterogeneous treatment effects: Causal Trees and Forests
•    Double Robust IPW
•    Applications

Learning Support:
•    Kozlowski and Evans (2019) “The geometry of Culture”, American Sociological Review.
•    Devlin, J., Chang, M.W., Lee, K., et al. (2019) BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding. Proceedings of NAACL-HLT 2019, Minneapolis, 2-7 June 2019, 4171-7186.
•    Chernozhukov, et al (2018), Double/debiased machine learning for treatment and structural parameters. The Econometrics Journal.
•    Michael Knaus, Slides 5, 6, 7 (2025);
•    Victor Chernozhukov, Hansen, Spindler, Kallus, Syrgkanis, (2026), Applied Causal Inference Powered by ML and AI. Chapters: 4, 9, 13, 15;
•    Mutlu Yuksel and Yigit Aydede, 2025, Causal Inference and Machine Learning: In Economics, Social, and Health Sciences. Chapters: 19, 21, 22, 23, 24;
•    S. Athey, & G. Imbens, Recursive partitioning for heterogeneous causal effects, Proc. Natl. Acad. Sci. U.S.A. 113 (27) 7353-7360, https://doi.org/10.1073/pnas.1510489113 (2016).

 

Register for this course