from https://www.kaggle.com/c/song-popularity-prediction
my first solo kaggle competition for predict song popularity. i use IterativeImputer as fillna value in data i use PCA 2 compomponent to feature engineer and speed the training i use Random Forest as tree-based that not affected by outlier and imbalance data
i get AUC score 0.51073 in private dashboard rank 453 from 535 team. it's ok for my first solo :)