000			01930nam a22002657a 4500
003			OSt
005			20220510005305.0
008			220510b |||||||| |||| 00| 0 eng d
020			_a9781108832373
040			_c0
082			_a005.758 GUO
092			_20
100			_aSong, Guo
245			_aEdge learning for distributed big data analytics _btheory, algorithms, and system design _cGuo Song, Zhihao Qu.
260			_aNew York _bCambridge University Press _c2021
300			_axi, 217 pages _c30 cm
504			_a Includes bibliographical references and index.
520			_aTraditionally, to develop these intelligent services and applications, big data are stored and processed in a centralized model. However, with the proliferation of edge devices and edge data, traditional centralized learning frameworks are required to upload all training data from different sources to a remote data server, which incurs significant communication overhead, service latency, as well as security and privacy issues. Therefore, it is urgent to shift model training and inference from the cloud to the edge, which is the essential idea of edge learning. Edge Learning is a fusion of big data, edge computing, and machine learning, and it is an enabling technology for edge intelligence. This book presents the basic knowledge of training machine learning models, key challenges and issues in edge learning, and comprehensive techniques from three aspects, i.e., fundamental theory, edge learning algorithms, and system design issues in edge learning. We believe that this book will stimulate fruitful discussions, inspire further research ideas, and attract researchers and developers from both academia and industry in this field
650			_aEdge computing
650			_aCOMPUTERS / Database Administration & Management
700			_aQu, Zhihao
843			_aPhotocopy
887			_2CamTech Library
942			_2ddc _cBK _n0
999			_c161 _d161