KDD Cup 2013 - author-paper identification challenge: second place team

Abstract

This paper describes our submission to the KDD Cup 2013 Track 1 Challenge: Author-Paper Indentification in the Microsoft Academic Search database. Our approach is based on Gradient Boosting Machine (GBM) of Friedman ([5]) and deep feature engineering. The method was second in the final standings with Mean Average Precision (MAP) of 0.98144, while the… (More)
DOI: 10.1145/2517288.2517291

1 Figure or Table

Topics

  • Presentations referencing similar topics