The Data Mining Forum
IMPORTANT: This is the old Data Mining forum.
I keep it online so that you can read the old messages.
Please post your new messages in the
new forum:
https://forum2.philippe-fournier-viger.com/index.php
Goto Topic:
Previous
•
Next
Goto:
Forum List
•
Message List
•
New Topic
•
Search
•
Log In
•
Print View
聚类算法的使用
Posted by:
qinqinzhou
Date: December 15, 2021 06:47PM
尊敬的philfv先生您好:
我在使用spmf库里的k-means聚类算法,数据是2436条67维的,拟采用欧氏距离聚成5类,但是聚类的效果不太好,聚类的结果文本也看不出来类间的区分,这是因为数据稀疏且维度太大了嘛?对于这样的数据聚类您有什么意见?非常感谢!
参数设置及可视化结果:
输入数据:
输出文本结果:
Options:
Reply
•
Quote
Re: 聚类算法的使用
Posted by:
qinqinzhou
Date: December 15, 2021 06:51PM
Options:
Reply
•
Quote
Re: 聚类算法的使用
Posted by:
webmasterphilfv
Date: December 23, 2021 09:05PM
Good afternoon,
I am sorry for the delay to answer. I have been very busy.
I think it is likely because the data is very sparse with so many dimensions.
In the visualization, the results may look strange but I think it is because we can only visualize two dimensions at a time.
By trying to look more in your data, maybe you can find some other reasons.
Best regards,
Philippe
Options:
Reply
•
Quote
Re: 聚类算法的使用
Posted by:
琴琴周
Date: December 24, 2021 11:50PM
Thank you very much!I am tring to reduce the dimensions,filling the gap values and reading papers,I think I would solve the problem.
Options:
Reply
•
Quote
This forum is powered by
Phorum
and provided by
P. Fournier-Viger
(© 2012).
Terms of use
.