New statistical method uses Gaussian processes for causal inference in time series

Researchers have developed a new statistical method for causal inference in interrupted time series designs, particularly useful when a treatment affects all units simultaneously. The approach uses Gaussian process regression to estimate counterfactuals by extrapolating from pre-treatment data, retaining functions consistent with the historical series and widening uncertainty intervals where extrapolation magnifies divergence. This method is demonstrated through simulations and an analysis of handgun purchases following the Supreme Court's Heller decision, with an accompanying R package named 'gpss' available for implementation. AI

IMPACT Introduces a novel statistical technique for causal inference, potentially improving analysis in fields utilizing time-series data.

RANK_REASON This is a research paper detailing a new statistical methodology. [lever_c_demoted from research: ic=1 ai=0.4]

Read on arXiv stat.ML →

• arXiv
• causal inference
• Gaussian process
• GPSS
• Heller decision
• R
• Supreme Court of the United States

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →