Tree-Based Models

Heard about this idea about Krish told me that he messed up his interview because he was specialized in this, which Sagen didn’t like?

Tree-Based Models > Neural Networks in competition??

• https://www.reddit.com/r/MachineLearning/comments/nxcbge/discussion_xgboost_is_the_way/
• https://towardsdatascience.com/when-and-why-tree-based-models-often-outperform-neural-networks-ceba9ecd0fd8

“Any Kaggler knows that XGBoost is by far the most popular choice for top-performing competition submissions. In essence, what puts tree-based methods and neural networks in the same category is that they approach problems by deconstructing them piece-by-piece, instead of finding one complex boundary that can separate the entire dataset, like SVM or Logistic Regression”.