Сравнение методов

	Машинное обучение с дополненной прерванной временной серией	Анализ причинно-следственного воздействия (Causal Impact Analysis)
Область	Причинно-следственный вывод	Причинно-следственный вывод
Семейство	Regression model	Regression model
Год появления	2014-2015	2015
Автор метода	Brodersen et al. (2015); Varian (2014) — foundational ML-for-causal-inference literature	Kay H. Brodersen, Fabian Gallusser, Jim Koehler, Nicolas Remy, Steven L. Scott (Google)
Тип	Quasi-experimental causal inference with ML counterfactual	Bayesian causal inference / counterfactual forecasting
Основополагающий источник	Brodersen, K. H., Gallusser, F., Koehler, J., Remy, N., & Scott, S. L. (2015). Inferring causal impact using Bayesian structural time-series models. Annals of Applied Statistics, 9(1), 247-274.	Brodersen, K. H., Gallusser, F., Koehler, J., Remy, N., & Scott, S. L. (2015). Inferring causal impact using Bayesian structural time-series models. Annals of Applied Statistics, 9(1), 247-274.
Другие названия	ML-ITS, ML-augmented ITS, machine learning ITS, causal ML interrupted time series	CausalImpact, BSTS causal inference, Bayesian causal impact, counterfactual time-series analysis
Связанные	6	5
Сводка	Machine Learning-Augmented Interrupted Time Series (ML-ITS) estimates the causal effect of a discrete intervention by training a machine learning model on pre-intervention time series data, projecting a counterfactual trajectory into the post-intervention period, and measuring the gap between observed and predicted outcomes. It extends classical ITS by replacing parametric trend assumptions with flexible ML estimators such as gradient boosting, random forests, or Bayesian structural time-series models.	Causal Impact Analysis, introduced by Brodersen et al. (2015) at Google, uses Bayesian structural time-series models to estimate what would have happened to an outcome had an intervention never occurred. By constructing a probabilistic counterfactual from pre-treatment data and control covariates, it quantifies point-in-time and cumulative treatment effects with full posterior uncertainty intervals.