Data Mining Tasks:Clustering

 

This one is for those friends who love to play with ‘DATA’. And frankly speaking playing with data is not a easy job as DATA is the at the beginning of the journey which ends at ‘Knowledge’.But a lot of efforts is required to discover the knowledge out of data.That’s what data mining tasks are all about.Lets discuss one important data mining task i.e. clustering here.

Definitions:

• the process of organizing objects into groups whose members are similar in one way or the other.
• A form of unsupervised learning, where we dont have examples demonstrating how the data should be grouped together.
• A method of data exploration i.e. a way of looking for patterns in data ,that are of interest.

so, a cluster is a collection of data objects which are similar among them but dissimilar to data objects belonging to other clusters.

Goal :

the goal of clustering is to determine the intrinsic grouping in a set of data.

Types:

•  Partitioning
• Hierarchical  

We will discuss hierarchical clustering technique in detail.    

Hierarchical clustering is further divided into two groups.They are:

• Agglomerative
• Divisive

In agglomerative hierarchical clustering ,following sequence of steps are followed.

• We start with every data item in a separate cluster.
• then we keep merging the most similar pair of clusters until we have one big   cluster left.

This is the bottom-up approach.

Good bye friends….