kmcuda:NVIDIA GPU CUDA上的大规模K-means和K-nn实施-源码

上传者: 42133861 | 上传时间: 2021-05-12 19:32:49 | 文件大小: 423KB | 文件类型: ZIP
使用NVIDIA CUDA的“阴阳” K均值和K-nn K均值的实现基于 。 尽管它引入了一些开销和许多对CUDA不利的条件子句,但与Lloyd算法相比,它仍然显示出1.6到2倍的加速。 K近邻采用相同的三角形不等式思想,并且需要预先计算的质心和聚类分配,类似于展平的球树。 sklearn KMeans KMeansRex KMeansRex OpenMP 塞班 克库达 kmcuda 2 GPU 速度 1倍 4.5倍 8.2倍 15.5倍 17.8倍 29.8倍 记忆 1倍 2倍 2倍 0.6倍 0.6倍 0.6倍 从技术上讲,该项目是一个共享库,可导出kmcuda.h定义的两个函数: kmeans_cuda和knn_cuda 。 它具有内置的Python3和R本机扩展支持,因此您可以from libKMCUDA import kmeans_cuda或dyn.load("libKMCUDA.so") 。 目录 K均值 该项目与其他项目之间的主要区别在于,kmcuda已针对内存消耗低和大量群集进行了优化。 例如,kmcuda可以将480个维度中的4M个样本分类为40000个

文件下载

资源详情

[{"title":"( 33 个子文件 423KB ) kmcuda:NVIDIA GPU CUDA上的大规模K-means和K-nn实施-源码","children":[{"title":"kmcuda-master","children":[{"title":".github","children":[{"title":"dco.yml <span style='color:#111;'> 25B </span>","children":null,"spread":false}],"spread":true},{"title":"DCO <span style='color:#111;'> 1.39KB </span>","children":null,"spread":false},{"title":"src","children":[{"title":"test.py <span style='color:#111;'> 29.10KB </span>","children":null,"spread":false},{"title":"tricks.cuh <span style='color:#111;'> 2.20KB </span>","children":null,"spread":false},{"title":"wrappers.h <span style='color:#111;'> 964B </span>","children":null,"spread":false},{"title":"CMakeLists.txt <span style='color:#111;'> 3.27KB </span>","children":null,"spread":false},{"title":"kmeans.cu <span style='color:#111;'> 46.53KB </span>","children":null,"spread":false},{"title":"cmake","children":[{"title":"FindR.cmake <span style='color:#111;'> 2.13KB </span>","children":null,"spread":false}],"spread":true},{"title":"test.R <span style='color:#111;'> 4.11KB </span>","children":null,"spread":false},{"title":"setup.py <span style='color:#111;'> 3.84KB </span>","children":null,"spread":false},{"title":"README.md <span style='color:#111;'> 12B </span>","children":null,"spread":false},{"title":"metric_abstraction.h <span style='color:#111;'> 9.14KB </span>","children":null,"spread":false},{"title":"knn.cu <span style='color:#111;'> 18.83KB </span>","children":null,"spread":false},{"title":"kmcuda.h <span style='color:#111;'> 8.67KB </span>","children":null,"spread":false},{"title":"fp_abstraction.h <span style='color:#111;'> 2.70KB </span>","children":null,"spread":false},{"title":"private.h <span style='color:#111;'> 15.06KB </span>","children":null,"spread":false},{"title":"python.cc <span style='color:#111;'> 22.14KB </span>","children":null,"spread":false},{"title":"transpose.cu <span style='color:#111;'> 3.99KB </span>","children":null,"spread":false},{"title":"kmcuda.cc <span style='color:#111;'> 26.94KB </span>","children":null,"spread":false},{"title":"r.cc <span style='color:#111;'> 14.85KB </span>","children":null,"spread":false}],"spread":false},{"title":"img","children":[{"title":"cls_angular.png <span style='color:#111;'> 37.86KB </span>","children":null,"spread":false},{"title":"kmeans_image.ipynb <span style='color:#111;'> 362.83KB </span>","children":null,"spread":false},{"title":"cls_euclidean.png <span style='color:#111;'> 26.22KB </span>","children":null,"spread":false},{"title":"sourced.png <span style='color:#111;'> 22.82KB </span>","children":null,"spread":false},{"title":"latex_angular.png <span style='color:#111;'> 3.21KB </span>","children":null,"spread":false}],"spread":true},{"title":"MAINTAINERS <span style='color:#111;'> 52B </span>","children":null,"spread":false},{"title":"doc","children":[{"title":"Doxyfile <span style='color:#111;'> 329B </span>","children":null,"spread":false}],"spread":true},{"title":"CONTRIBUTING.md <span style='color:#111;'> 2.52KB </span>","children":null,"spread":false},{"title":".travis.yml <span style='color:#111;'> 845B </span>","children":null,"spread":false},{"title":"LICENSE.md <span style='color:#111;'> 10.25KB </span>","children":null,"spread":false},{"title":"README.md <span style='color:#111;'> 28.84KB </span>","children":null,"spread":false},{"title":"CODE_OF_CONDUCT.md <span style='color:#111;'> 3.15KB </span>","children":null,"spread":false},{"title":".gitignore <span style='color:#111;'> 388B </span>","children":null,"spread":false}],"spread":false}],"spread":true}]

评论信息

免责申明

【只为小站】的资源来自网友分享,仅供学习研究,请务必在下载后24小时内给予删除,不得用于其他任何用途,否则后果自负。基于互联网的特殊性,【只为小站】 无法对用户传输的作品、信息、内容的权属或合法性、合规性、真实性、科学性、完整权、有效性等进行实质审查;无论 【只为小站】 经营者是否已进行审查,用户均应自行承担因其传输的作品、信息、内容而可能或已经产生的侵权或权属纠纷等法律责任。
本站所有资源不代表本站的观点或立场,基于网友分享,根据中国法律《信息网络传播权保护条例》第二十二条之规定,若资源存在侵权或相关问题请联系本站客服人员,zhiweidada#qq.com,请把#换成@,本站将给予最大的支持与配合,做到及时反馈和处理。关于更多版权及免责申明参见 版权及免责申明