CatBoost is specifically designed to handle categorical features efficiently. How does CatBoost process categorical features during training? a$)$one$-$hot encoding b$)$CatBoost automatically handles c$)$Target Encoding d$)$Label Encoding