TY - GEN
T1 - EdgePS
T2 - 14th IEEE International Conference on Cloud Computing, CLOUD 2021
AU - Zhao, Yangming
AU - Hou, Yunfei
AU - Qiao, Chunming
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021/9
Y1 - 2021/9
N2 - In this paper, we propose EdgePS, an advanced parameter server approach for distributed machine learning in edge computing scenarios. Different from the Conventional Parameter Server (CPS) approach, which performs parameter aggregation after every local training epoch, EdgePS synchronizes the parameters of all workers only when the local training cannot improve the global model performance. We first analyze how the local training will impact the performance of the global model, and then design algorithms to determine when the best time is to perform the parameter aggregation. Both real testbed experiments and extensive large scale simulations demonstrate that EdgePS can train a practical machine learning model, e.g., VGG-16, with up to 59.28% less time compared with the CPS approach. With the same training time, EdgePS can improve model accuracy by up to 30.19 % compared with the state-of-The-Art distributed machine learning algorithm designed for edge computing scenarios.
AB - In this paper, we propose EdgePS, an advanced parameter server approach for distributed machine learning in edge computing scenarios. Different from the Conventional Parameter Server (CPS) approach, which performs parameter aggregation after every local training epoch, EdgePS synchronizes the parameters of all workers only when the local training cannot improve the global model performance. We first analyze how the local training will impact the performance of the global model, and then design algorithms to determine when the best time is to perform the parameter aggregation. Both real testbed experiments and extensive large scale simulations demonstrate that EdgePS can train a practical machine learning model, e.g., VGG-16, with up to 59.28% less time compared with the CPS approach. With the same training time, EdgePS can improve model accuracy by up to 30.19 % compared with the state-of-The-Art distributed machine learning algorithm designed for edge computing scenarios.
KW - n/a
UR - https://www.scopus.com/pages/publications/85119321731
U2 - 10.1109/CLOUD53861.2021.00035
DO - 10.1109/CLOUD53861.2021.00035
M3 - Conference contribution
AN - SCOPUS:85119321731
T3 - IEEE International Conference on Cloud Computing, CLOUD
SP - 217
EP - 227
BT - Proceedings - 2021 IEEE 14th International Conference on Cloud Computing, CLOUD 2021
A2 - Ardagna, Claudio Agostino
A2 - Chang, Carl K.
A2 - Daminai, Ernesto
A2 - Ranjan, Rajiv
A2 - Wang, Zhongjie
A2 - Ward, Robert
A2 - Zhang, Jia
A2 - Zhang, Wensheng
PB - IEEE Computer Society
Y2 - 5 September 2021 through 11 September 2021
ER -