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
How to use clustering for a course project (DBScan and KMeans)?
Posted by:
Hakrith
Date: April 21, 2011 05:51AM
Hello,
I'm a B.SC. student in computer science doing a course on data mining this semester. I have to do a data mining project.
I have found some source code for K-Means and DBScan and I would like to do a project with these algorithms.
Please give me some ideas about how to use K-Means and DBScan in a concrete application. I have tought about clustering products for a computer store but i'm not sure if it is interesting. I'm also not sure how to make the clusters not too large.
Thanks
Hakrith
I'm a B.SC. student in computer science doing a course on data mining this semester. I have to do a data mining project.
I have found some source code for K-Means and DBScan and I would like to do a project with these algorithms.
Please give me some ideas about how to use K-Means and DBScan in a concrete application. I have tought about clustering products for a computer store but i'm not sure if it is interesting. I'm also not sure how to make the clusters not too large.
Thanks
Hakrith
Re: How to use clustering for a course project (DBScan and KMeans)?
Posted by:
Symboldrive
Date: April 26, 2011 11:37PM
Hi Hakrith
DBScan and KMEANS have a lot of applications. I give you some examples.
It can be used for example to cluster customers of a bank that have similar characteristics (type of accounts, sex, age, etc.). Then the bank could use this to take some decision about similar customers. For example, it could be used to assess if you are likely to pay back money that you have borrowed from the bank.
Clustering could be also used in e-learning to dynamically create teams for learning activities.
For the source code of these algorithms, I think that you can search Google. K-Means and DBScan are very famous algorithms so it should be easy to find some source code if you don't want to program these algorithms by yourself.
But I encourage you to try to program these algorithms by yourself since they are easy to implement.
There is a pretty good description of DBScan on Wikipedia :
http://en.wikipedia.org/wiki/DBSCAN
And KMeans :
http://en.wikipedia.org/wiki/K-means_clustering
Good luck!
DBScan and KMEANS have a lot of applications. I give you some examples.
It can be used for example to cluster customers of a bank that have similar characteristics (type of accounts, sex, age, etc.). Then the bank could use this to take some decision about similar customers. For example, it could be used to assess if you are likely to pay back money that you have borrowed from the bank.
Clustering could be also used in e-learning to dynamically create teams for learning activities.
For the source code of these algorithms, I think that you can search Google. K-Means and DBScan are very famous algorithms so it should be easy to find some source code if you don't want to program these algorithms by yourself.
But I encourage you to try to program these algorithms by yourself since they are easy to implement.
There is a pretty good description of DBScan on Wikipedia :
http://en.wikipedia.org/wiki/DBSCAN
And KMeans :
http://en.wikipedia.org/wiki/K-means_clustering
Good luck!