Building models with SAS Enterprise Miner, SAS Factory Miner, SAS Visual Data Mining and Machine Learning or just with programming

StatExplore Node Cluster Analysis

Reply
New Contributor
Posts: 2

StatExplore Node Cluster Analysis

Hello,

I have used a StatExpore node to find out the mean, standard deviation etc of a set of data.

Im trying to work out which field is the most consistent, I'm thinking its the no_inhab field as the mean is closest to the standard deviation, is that correct?

Also any ideas on other interesting features of the data would be helpful? Don't particularly want a straight answer, more an explanation of what to explore.

Thanks in advance


Screen Shot 2014-11-18 at 23.41.32.png
Trusted Advisor
Posts: 1,231

Re: StatExplore Node Cluster Analysis

Hi,

In order to find more consistent variable, usually coefficient of variation is considered, no_inhab field has mean and standard deviation very close but this variable is highly skewed and may contain outliers, as you can see its maximum value which is not consistent with the mean.

Thanks,

Naeem

New Contributor
Posts: 2

Re: StatExplore Node Cluster Analysis

Thanks.

How does the coefficient of the variable help find a consistent variable? What does it mean if the coefficient is low or high?

Trusted Advisor
Posts: 1,231

Re: StatExplore Node Cluster Analysis

Cofficient of varition expresses the standard deviation as percentage of the arithmetic mean. Low value means more consistent.

SAS Employee
Posts: 26

Re: StatExplore Node Cluster Analysis

Hello SamBan - Perhaps you can tell us a little more about the nature of your problem here to help us guide you on what to explore.  What exactly are you looking for when you say you are looking for the field/variable which is most consistent?

Thanks,

Brett

Ask a Question
Discussion stats
  • 4 replies
  • 392 views
  • 3 likes
  • 3 in conversation