Empirical work in economics is typically concerned with causal inference and hypothesis testing, whereas machine learning is centered around prediction. There is, however, a clear intersection between objectives when it comes to forecasting in economics and finance. Consequently, there has been increasing interest in using methods from machine learning to produce and evaluate economic forecasts.
In Chapter 2, we discussed Coulombe et al. (2019), which evaluated the usefulness of machine learning for time series econometrics. They identified non-linear models, regularization, ...
