CUDA on GPU kMeans
2016-08-23
0 0 0
4.0
Other
Earn points
Application background
CUDA on the kMeans to achieve GPU, a cluster analysis.We often come into contact with the clustering analysis. It is a common practice to extract N features, and then put them together to form a N vector, so that a mapping from the original data set to a N dimensional vector space is needed.Can also be applied to image segmentation, etc..Key Technology
CUDA GPU implementation of kMeans, parallel computing, data division and other related technologies.The steps of the algorithm are as follows:1, from the D K random elements, as the K of the cluster of their respective centers.2, calculated the remaining elements to the centers K cluster dissimilarity, these elements will be classified to the lowest cluster dissimilarity.3. According to the results of clustering, the K cluster is calculated. The calculation method is the arithmetic mean of all the elements in the cluster.4, the D of all elements in accordance with the new cec
cuda
GPU
KMeans
实现
Related Source Codes
PClite
0
0
no vote
generic embeded FTP Client
0
0
no vote
UDP Test Demo
0
0
no vote
TMS320F28335 DSP transmits data to EEPROM through
0
0
no vote
TMS320F28335 CAN communication source code
0
0
no vote
No comment