The dynamic generalized covariance measure for conditional independence testing with nonstationary time series
The episode discusses a paper titled "The Dynamic Generalized Covariance Measure for Conditional Independence Testing with Nonstationary Time Series." The hosts explain that this method allows researchers to test conditional independence between variables even when the underlying data is nonstationary and changing over time. They detail how the test uses time-varying regressions and cumulative error sums to detect fleeting connections in real-world, messy data.