Simulated Data, Real Learnings: Part 1

Testing machine learning approaches with simulation

distributions of model estimated coefficients on simulated data — image by author

Simulation is a powerful tool in the data science tool box. This is the first part of a multi-part series that discusses various ways that simulation can be useful in data science and machine learning. In this article, we will cover how simulation can be used to test machine learning approaches.