Categorical Embeddings with CatBoost

Investigating whether using categorical embeddings from a fast.ai neural network can improve the performance of CatBoost in a classification task

CatBoost is an open-sourced machine learning algorithm from Yandex. Its name is derived from the words Category Boosting. As you might expect, one of the library’s main advantages is handling categorical data in a more intuitive way than other decision tree algorithms. It also performs favourably compared to other decision tree algorithms. …

catboost embedding fastai machine learning regression