Classification with a Tabular Vector Borne Disease Dataset

• Classification with a Tabular Vector Borne Disease Dataset | Kaggle
• Got first place in the DataFeeling leaderboard
• Won some merch Telegram: @data_science_winners

Task and solution

• The task involves multi-class classification on tabular data.
• The features represent symptoms of horses.
• The evaluation metric used is MPA@3.
• The solution is a grid-searched CatBoost model with loss_function="MultiClass", without any feature engineering.
• The final predictions were generated by selecting the top 3 classes based on their probabilities.