/kaggle-Avazu-Clickthrough-Rate-Prediction

Python kernels for exploratory data analysis, feature engineering, modeling and evaluation, using two different approaches: gradient boosting machines with LightGBM, and logistic regression.

Primary LanguageJupyter NotebookMIT LicenseMIT

kaggle_avazu_clickthrough_rate

Kaggle Avazu clickthrough rate prediction competition https://www.kaggle.com/c/avazu-ctr-prediction Python kernels for exploratory data analysis, feature engineering, modeling and evaluation

The folder sgd contains stochastic gradient descent solution

The folder sklearn contains gradient boosting machines solution, using lightgbm