Author:
Yang Can,Li Yin,Cheng Fenhua
Abstract
Abstract
We accelerate basic k-Means algorithm using CUDA GPU, a new programming model by NVIDIA, and experiment data shows we achieve a maximum speedup of 67.752, while other teams claim 20 to 40. Also we find that the basic k-Means algorithm is most sensitive to the cluster size k, and less to the datasets size b and least to the dimension d. In addition, we find the CUDA shared memory improves the performance, but also depends on which factor we scale.
Reference12 articles.
1. An Efficient k-Means Clustering Algorithm: Analysis and Implementation;Kanungo;IEEE Transactions on Pattern Analysis & Machine Intelligence,2002
2. An Efficient Approximation to the k-Means Clustering for Massive Data;Capóa;Knowledge-Based Systems,2016
3. The planar k-means problem is NP-hard;Mahajan,2012
4. Scalable k-Means Clustering via Lightweight Coresets;Bachem,2017
Cited by
3 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. Parametric Pipelined k-Means Implementation for Hyperspectral Processing on Spacecraft Embedded FPGA;IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing;2024
2. Fabric Image Layering Based on Parallel K-means;2022 IEEE 24th Int Conf on High Performance Computing & Communications; 8th Int Conf on Data Science & Systems; 20th Int Conf on Smart City; 8th Int Conf on Dependability in Sensor, Cloud & Big Data Systems & Application (HPCC/DSS/SmartCity/DependSys);2022-12
3. Large scale K-means clustering using GPUs;Data Mining and Knowledge Discovery;2022-10-18