GPU上的决策树构造：无处不在的并行计算方法.pdf

Computing (2014) 96:403–413 DOI 10.1007/s00607-013-0343-z Decision tree construction on GPU: ubiquitous parallel computing approach Aziz Nasridinov · Yangsun Lee · Young-Ho Park Received: 11 April 2013 / Accepted: 8 August 2013 / Published online: 24 August 2013 © Springer-Verlag Wien 2013 Abstract General Purpose Graphic Processing Unit (GPGPU) computing with CUDA has been effectively used in scientific applications, where huge accelerations have been achieved. However, while today’s traditional GPGPU can reduce the exe- cution time of parallel code by m