FOOD SCIENCE ›› 2005, Vol. 26 ›› Issue (6): 109-112.
Previous Articles Next Articles
LIU Jian-Xue, LI Shou-Jun
Online:
Published:
Abstract: Cosine similarity is a good reflection of vector’s similarity, and it also reflect the variations of vector’s elements. Using cosine similarity as a scale of clustering could better tell how many variations contained in the vectors and could find some eigenvectors to substitute all vectors in evaluation. The eigenvectors found by clustering could be used in regression. This article described a means of dynamically clustering with Cosine Similarity, forty-nine rice samples’ infrared spectra had been used to test the means of clustering, and 9 independent clusters were identified.
Key words: cosine similarity, eigenvector, dynamical clustering, near infrared spectra
LIU Jian-Xue, LI Shou-Jun. Dynamical Clustering by Cosine Similarity Algorithm[J]. FOOD SCIENCE, 2005, 26(6): 109-112.
0 / / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: https://www.spkx.net.cn/EN/
https://www.spkx.net.cn/EN/Y2005/V26/I6/109