Skip to main content

😺 CatBoost

Created by Yandex, CatBoost is the master of Categorical Data (text labels).

🏷️ The Categorical Nightmare​

Usually, you have to convert text like "New York" into numbers. CatBoost handles text categories out of the box!

🐍 Python Implementation​

# pip install catboost
from catboost import CatBoostClassifier

# Notice X has TEXT directly in it!
X = [["New York", 10], ["Paris", 20], ["Tokyo", 30], ["New York", 15]]
y = [1, 0, 0, 1]

# We tell CatBoost which column has the text (index 0)
model = CatBoostClassifier(iterations=10, cat_features=[0], verbose=0)
model.fit(X, y)

print("Prediction:", model.predict([["Paris", 10]]))