TY - GEN
T1 - Filter pruning of Convolutional Neural Networks for text classification
T2 - 2018 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2018
AU - Yoon, Hong Jun
AU - Robinson, Sarah
AU - Christian, J. Blair
AU - Qiu, John X.
AU - Tourassi, Georgia D.
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/4/6
Y1 - 2018/4/6
N2 - Convolutional Neural Networks (CNN) have recently demonstrated effective performance in many Natural Language Processing tasks. In this study, we explore a novel approach for pruning a CNN's convolution filters using our new data-driven utility score. We have applied this technique to an information extraction task of classifying a dataset of cancer pathology reports by cancer type, a highly imbalanced dataset. Compared to standard CNN training, our new algorithm resulted in a nearly.07 increase in the micro-averaged F1-score and a strong.22 increase in the macro-averaged F1-score using a model with nearly a third fewer network weights. We show how directly utilizing a network's interpretation of data can result in strong performance gains, particularly with severely imbalanced datasets.
AB - Convolutional Neural Networks (CNN) have recently demonstrated effective performance in many Natural Language Processing tasks. In this study, we explore a novel approach for pruning a CNN's convolution filters using our new data-driven utility score. We have applied this technique to an information extraction task of classifying a dataset of cancer pathology reports by cancer type, a highly imbalanced dataset. Compared to standard CNN training, our new algorithm resulted in a nearly.07 increase in the micro-averaged F1-score and a strong.22 increase in the macro-averaged F1-score using a model with nearly a third fewer network weights. We show how directly utilizing a network's interpretation of data can result in strong performance gains, particularly with severely imbalanced datasets.
UR - http://www.scopus.com/inward/record.url?scp=85050879139&partnerID=8YFLogxK
U2 - 10.1109/BHI.2018.8333439
DO - 10.1109/BHI.2018.8333439
M3 - Conference contribution
AN - SCOPUS:85050879139
T3 - 2018 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2018
SP - 345
EP - 348
BT - 2018 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2018
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 4 March 2018 through 7 March 2018
ER -
