DBSCAN(Density-Based Spatial Clustering of Applications with Noise)是一个比较有代表性的基于密度的聚类算法。与划分和层次聚类方法不同，它将簇定义为密度相连的点的最大集合，能够把具有足够高密度的区域划分为簇，并可在噪声的空间数据库中发现任意形状的聚类。
SHOW FULL COLUMNS FROM `jrk_downrecords` [ RunTime:0.001347s ]
SELECT `a`.`aid`,`a`.`title`,`a`.`create_time`,`m`.`username` FROM `jrk_downrecords` `a` INNER JOIN `jrk_member` `m` ON `a`.`uid`=`m`.`id` WHERE `a`.`status` = 1 GROUP BY `a`.`aid` ORDER BY `a`.`create_time` DESC LIMIT 10 [ RunTime:0.091622s ]
SHOW FULL COLUMNS FROM `jrk_tagrecords` [ RunTime:0.001229s ]
SELECT * FROM `jrk_tagrecords` WHERE `status` = 1 ORDER BY `num` DESC LIMIT 20 [ RunTime:0.001968s ]
SHOW FULL COLUMNS FROM `jrk_member` [ RunTime:0.001230s ]
SELECT `id`,`username`,`userhead`,`usertime` FROM `jrk_member` WHERE `status` = 1 ORDER BY `usertime` DESC LIMIT 10 [ RunTime:0.005028s ]
SHOW FULL COLUMNS FROM `jrk_searchrecords` [ RunTime:0.001058s ]
SELECT * FROM `jrk_searchrecords` WHERE `status` = 1 ORDER BY `num` DESC LIMIT 5 [ RunTime:0.003215s ]
SELECT aid,title,count(aid) as c FROM `jrk_downrecords` GROUP BY `aid` ORDER BY `c` DESC LIMIT 10 [ RunTime:0.014816s ]
SHOW FULL COLUMNS FROM `jrk_articles` [ RunTime:0.001357s ]
UPDATE `jrk_articles` SET `hits` = 1 WHERE `id` = 323068 [ RunTime:0.001401s ]
