Patrick Crenshaw
Re-searching Hyperparameters for Training Boosted Tree Models
Introduction While deep neural networks have state-of-the-art performance in many tasks, boosted tree models still often outperform deep neural networks on tabular data. This largely seems to be the c[…]