Clustering by fast search and find of density peaks代碼詳解，clusteringpeaks

Clustering by fast search and find of density peaks代碼詳解，clusteringpeaks


發表在2014年Science上的聚類方法Clustering by fast search and find of density peaks，不在這贅述介紹了，其matlab代碼詳細注釋下面給出，如果想看該方法的詳細解釋說明，推薦在閱讀原文的基礎上，參考這位博主的博文

http://blog.csdn.net/itplus/article/details/38926837

本文著作權  轉載請註明出處

http://blog.csdn.net/mxlmhgzw

無法看清的，請見我部落格

http://blog.sina.com.cn/s/blog_6ca0f5eb0102vdt8.html

如果運行代碼報錯，

     查出Y1 = mdscale(dist, 2, 'criterion','metricsstress');這句報錯

??? Error using ==> mdscale>metricStressCrit at 686

Points in the configuration have co-located.  Try a different starting point, or use a different criterion.

 

Error in ==> mdscale>MDS at 498

    [stress,grad] = feval_r(stressFun, Y, disparities, weights);

 

Error in ==> mdscale at 404

            [Y,stress,iter] = MDS(Y0,dissimilarities,weights,stressFun,metric,weighted,options);

 

Error in ==> cluster_dp at 181

可以修改最後一個參數，把它改成'metricsstress' 或 'sammon' 或 'strain'