Randomized Feature Engineering as a Fast and Accurate Alternative to Kernel Methods

Feature engineering has found increasing interest in recent years because of its ability to improve the effectiveness of various machine learning models. Although tailored feature engineering methods