LAE

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: kursattaskin

Hi everyone
I want to introduce you with three simple outlier detection techniques that are widely used

1- Three-Sigma Edit (ESD) which is actually getting the tails of a Gaussian distribution
2- Hampel Identifier which is getting the outliers according to median and MAD (Median Absolute Deviation)
3- Box Plot Outlier which is using box plot methodology

I will share when I develop other outlier detection algorithms depth-based approaches, deviation based approaches, distance based approaches etc.
I will try to do my best with my little free time

hope this can help someone else
any comment will be appreciatedOutlier_Detection_example.brg

kind regards,

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: ltolleson

Nice work... Thanks for sharing. You might also want to check out the Lavastorm Analytic Library (LAL1). We have some additional analytic nodes in this library including "R Analytics" and "Enhanced Analytics" nodes.

You can download the libraries and additional information at http://www.lavastorm.com/softwaredownloadsandtrials.

Thanks,
Larry

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: Marwa.Abbes

Hey!


Thank you for sharing!
Can we get the input file you used: Sale Record? It will give us a better understanding of the different techniques

Thanks.