Orange Forum • View topic - Thanks to Orange, I discover a hidden pattern

Thanks to Orange, I discover a hidden pattern

A place to ask questions about methods in Orange and how they are used and other general support.

Thanks to Orange, I discover a hidden pattern

Postby Preneur_d_ames » Fri Nov 20, 2009 13:59

I will present a technical paper to an international conference in 2 weeks. It's about practices in data handling among some governement agencies in a particular region of the world.

I used Orange to mine behaviours and practices patterns. It was extremely helpful.
I used scatterplot on k-means clustering, MDS, HCA on distance of attributes and examples and distance map. Very simple analysis.

However, I got amazing results, I could for example be able to categorize the practices regarding risks.
I found that the implementation of data management systems has no impact on the level of risk showing they are not considering risk mitigation and prevention as a part of their policy. The clustering was as well extremely interesting and confirmed what my 'guts' were telling me about the typology of these agencies.

So congrats for this very easy-to-use but extensible tool.

Return to Questions & Support



cron