METHODS FOR ASSESSING CAUSAL EFFECTS BASED ON MACHINE LEARNING
Authors:
1.
FGBOU VO «Voronezhskiy gosudarstvennyy lesotehnicheskiy universitet imeni G.F. Morozova»
2.
Voronezh State University of Forestry and Technologies named after G.F. Morozov
Type:
Сonference article
Published:
25.06.2025
Subject area:
004.04
Language:
Russian
Keywords:
machine learning, Uplift, Gelman's method, T-Learner, class transformation method, treatment, Uplift - modeling
Abstract (English):
The article provides a comparative analysis of modern methods for assessing cause-and-effect effects based on machine learning: T-Learner, the Gelman method and the class transformation method, which help to evaluate heterogeneous effects of exposure in observational samples. The results obtained are based on an experimental study based on real data.
The article provides a comparative analysis of modern methods for assessing cause-and-effect effects based on machine learning: T-Learner, the Gelman method and the class transformation method, which help to evaluate heterogeneous effects of exposure in observational samples. The results obtained are based on an experimental study based on real data.
Keywords:
machine learning, Uplift, Gelman's method, T-Learner, class transformation method, treatment, Uplift - modeling
machine learning, Uplift, Gelman's method, T-Learner, class transformation method, treatment, Uplift - modeling
1. Pearl, J. Causal Inference in Statistics: A Primer / J. Pearl, M. Glymour, N.P. Jewell. – Wiley, 2016. – 160 p.
2. Imbens, G.W. Causal Inference for Statistics, Social, and Biomedical Sciences / G.W. Imbens, D.B. Rubin. – Cambridge University Press, 2015. – 640 p.
3. Bayesian Data Analysis / A. Gelman, J.B. Carlin, H.S. Stern [et al.]. –3rd ed. – Chapman & Hall/CRC, 2013. – 675 p.
4. Gutierrez, P. Uplift Modeling: From Causal Inference to Personalization. – URL: https://arxiv.org/abs/1705.08492 (data obrascheniya: 01.04.2025).
5. CausalML Documentation. – URL: https://causalml.readthedocs.io/(data obrascheniya: 01.04.2025).
6. Scikit-uplift Library. – URL: https://scikit-uplift.readthedocs.io/ (data obrascheniya: 01.04.2025).
