I've been teaching tabular data modeling to many clients. My usual model of choice these days is XGBoost. Thought, we can often go back to a strong linear model

I've been teaching tabular data modeling to many clients. My usual model of choice these days is XGBoost. Thought, we can often go back to a strong linear model based on our insights from XGBoost.