Summary of A Practical Approach to Causal Inference Over Time, by Martina Cinquini et al.
A Practical Approach to Causal Inference over Time
by Martina Cinquini, Isacco Beretta, Salvatore Ruggieri, Isabel Valera
First submitted to arxiv on: 14 Oct 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Machine Learning (cs.LG); Numerical Analysis (math.NA)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper presents a novel approach to estimating the causal effect of an intervention over time on a dynamic system. It formally defines causal interventions and their effects on discrete-time stochastic processes (DSPs) and shows how the equilibrium states of a DSP, both before and after an intervention, can be captured by a structural causal model (SCM). The authors then provide an explicit mapping from vector autoregressive models (VARs), commonly used in econometrics, to linear SCMs. This allows for causal inference over time from observational time series data using the proposed causal VAR framework. Experiments on synthetic and real-world datasets demonstrate strong performance in terms of observational forecasting and accurate estimation of the causal effect of interventions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper helps us understand how to figure out what happens when we do something new on a dynamic system, like a weather forecast model. It shows that if we can define exactly what this new thing does, and how it changes over time, we can use special models called structural causal models (SCMs) to make predictions about what will happen next. The authors also show how to connect these SCMs to another kind of model, vector autoregressive models (VARs), which are commonly used in economics. This lets us do something new and important – figure out the cause of changes we see in the data. The results look promising, with good accuracy on real-world datasets. |
Keywords
» Artificial intelligence » Autoregressive » Inference » Time series
