Guest
Daniel Savenkov tells his solution of Kaggle Mercedes-Benz Greener Manufacturing competition. In this competition, Daimler is challenging Kagglers to tackle the curse of dimensionality and reduce the time that cars spend on the test bench. From this video you learn: - How to detect competitions with great leaderboard shuffle - Why need you to analyze carefully public scripts before using them - How to build cross-validation with significance evaluating and outlier handling - Can predictions by clusters be better than usual - Why people take part in doubtful competition Slides (english): Links: - Author’s kaggle profile: - Author’s linkedin profile: - Solution at kaggle forum: #202443 - Solution’s code at github: - Stacked and Averaged models public script: - 10th place public script: - Cross-validation implementation: - Objects clustering: - 2nd place approach: - 1st place approach:
