Title :
Parsimonious Naive Bayes
Author_Institution :
Orange Labs., Lannion, France
Abstract :
We describe our submission to the AAIA´14 Data Mining Competition, where the objective was to reach good predictive performance on text mining classification problems while using a small number of variables. Our submission was ranked 6th, less than 1% behind the winner. We also present an empirical study on the trade-off between parsimony of the representation and accuracy, and show how good performance can be obtained quickly and efficiently.
Keywords :
Bayes methods; data mining; pattern classification; parsimonious naive Bayes; text mining classification problems; trade-off; Accuracy; Bayes methods; Data mining; Input variables; Lead; Niobium; Optimization;
Conference_Titel :
Computer Science and Information Systems (FedCSIS), 2014 Federated Conference on
Conference_Location :
Warsaw