TY - JOUR
T1 - Structured Bayesian Compression for Deep Models in Mobile-Enabled Devices for Connected Healthcare
AU - Chen, Sijia
AU - Song, Bin
AU - Du, Xiaojiang
AU - Guizani, Nadra
N1 - Publisher Copyright:
© 1986-2012 IEEE.
PY - 2020/3/1
Y1 - 2020/3/1
N2 - Deep models, typically deep neural networks, have millions of parameters, analyze medical data accurately, yet in a time-consuming method. However, energy cost effectiveness and computational efficiency are important for prerequisites developing and deploying mobile-enabled devices, the mainstream trend in connected healthcare. Therefore, deep models' compression has become a problem of great significance for real-time health services. In this article, we first emphasize the use of Bayesian learning for model sparsity, effectively reducing the number of parameters while maintaining model performance. Specifically, with sparsity inducing priors, large parts of the network can be pruned with a simple retraining of arbitrary datasets. Then, we propose a novel structured Bayesian compression architecture by adaptively learning both group sparse and block sparse while also designing sparse-oriented mixture priors to improve the expandability of the compression model. Experimental results from both simulated datasets (MNIST) as well as practical medical datasets (Histopathologic Cancer) demonstrate the effectiveness and good performance of our framework on deep model compression.
AB - Deep models, typically deep neural networks, have millions of parameters, analyze medical data accurately, yet in a time-consuming method. However, energy cost effectiveness and computational efficiency are important for prerequisites developing and deploying mobile-enabled devices, the mainstream trend in connected healthcare. Therefore, deep models' compression has become a problem of great significance for real-time health services. In this article, we first emphasize the use of Bayesian learning for model sparsity, effectively reducing the number of parameters while maintaining model performance. Specifically, with sparsity inducing priors, large parts of the network can be pruned with a simple retraining of arbitrary datasets. Then, we propose a novel structured Bayesian compression architecture by adaptively learning both group sparse and block sparse while also designing sparse-oriented mixture priors to improve the expandability of the compression model. Experimental results from both simulated datasets (MNIST) as well as practical medical datasets (Histopathologic Cancer) demonstrate the effectiveness and good performance of our framework on deep model compression.
UR - http://www.scopus.com/inward/record.url?scp=85071856293&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85071856293&partnerID=8YFLogxK
U2 - 10.1109/MNET.001.1900204
DO - 10.1109/MNET.001.1900204
M3 - Article
AN - SCOPUS:85071856293
SN - 0890-8044
VL - 34
SP - 142
EP - 149
JO - IEEE Network
JF - IEEE Network
IS - 2
M1 - 8823872
ER -