The methods

---

Constraint-based methods are a major approach in causal discovery,aiming to infer causal structures from observational data by analyzing conditional independence relationships.These methods rely on the Causal Markov Condition and d-separation to infer the underlying Directed Acyclic Graph(DAG)

Score-based methods for causal discovery search for the best causal structure by assigning a score to different candidate graphs and optimizing for the highest score. Unlike constraint-based methods, score-based methods use a global optimization approach to find the best Directed Acyclic Graph(DAG)

FCMs are a powerful framework for causal discovery ,repesenting how variables are generated in a system through explicit functional relationships.Unlike constraint-based and score-based methods, FCMs assume a specific functional form for causal mechanisms, allowing for more precise inference of causal relationships.

Granger causality is a statistical hypothesis test used to determine whether one time series can predict another. it is widely used in econometrics , neuroscience, and time series analysis to infer temporal causal relationships between variable.

It combines traditional causal inference methods with advanced data-driven techniques to learn the causal structure of a system from observational data.Machine Learning methods can be leveraged to automate ,optimize, and expand causal discovery,offering improvements over traditional statistical techniques, especially when dealing with large,high-dimensional datasets or complex non-linear relationships.