Data preprocessing for clustering: survey

The University of Jena (Germany) is currently working on preprocessing for clustering. It is part of a bigger project (WhiBo). They have prepared a survey that takes around 10 minutes to fill. Here is a brief explanation about their survey:

Although preprocessing as a first step in clustering is important, little effort is done in systematizing data preprocessing. That might cause problems for users of data mining software because they are not sure which algorithm they should apply in which situation.

Therefore we conducted a study to ask software users and project leaders in the data mining field about how they decide about what algorithms should be used and what are their experiences in solving preprocessing problems.

You can access the online questionnaire here. The deadline is December 6th.


Recommended Reading

Comments Icon2 comments found on “Data preprocessing for clustering: survey

  1. data preprocessing step for clustering can be done with PCA,FUZZY C_means and EM techniques

Comments are closed.