Back

Speaker "Daniel Hansen" Details Back

 

Topic

Text Classification

Abstract

Continuous vector space word representations are leveraged to significantly improve the performance of supervised text classification. Comparisons are made with other standard text classification approaches including a deep network pre-trained with Autoencoder layers. We find that our approach beats the others in terms of accuracy as well its ability to effectively learn from very small training sets.

Profile

Data Scientist at Personagraph