'Cluster Analysis in Weka, with the Bank dataset

I am using the bank dataset for cluster analysis in Weka. The dataset is here: bank-k.arff

However, the dataset appears to be not ideal for cluster analysis. When using algorithms like EM, k-means, and cobweb, the result does not show trends. Classes to cluster evaluations yield unhelpful results. For instance, when using the region as the class:

bank cluster weka

Most of the graphs are like this, which does not yield helpful results. In contrast, the iris dataset does return distinguishing results:

Iris in Weka

Am I doing it incorrectly? I am new in this field of study. Please give some advice if possible.

Think you in advance!

data-sciencecluster-analysisweka


Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source

Related Questions