On Thu, 2014-01-16 at 10:30 -0600, John McKown wrote: > And my even more OT digression: an article on not using "standard > deviation" as a measure > http://www.edge.org/response-detail/25401
I saw this on HN yesterday, where the comments were generally unkind. Statistics is not my forte but I've found STD useful for detecting outliers - exactly what we need for the email case we were discussing. -- David Andrews A. Duda & Sons, Inc. [email protected] ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
