Hello guys,

i am trying to profile a customer based on his transaction activity
history. This is done so that i can compare the profile (average
activity) against the current activity and determine if the current
activity falls within the limits of the profile. If it doesnt, I say
it is suspicious.

But if the customers were involved in suspicious activity in their
past, it skews the profile in such a way that even their profile is
suspicious. So there is no way for me to determine if their current
activity is suspicious.

So I was wondering if there is a way to determine (statistically)
their average activity (somehow excluding the suspicious)

Regarding your questions
1. I cannot eliminate them manually because i wouldnt know what to
eliminate.
2. the sample may range from thousands to 100 thousands. sometimes
millions. It is the transaction history over the past year(s) of a
customer.

-sat

[EMAIL PROTECTED] (Dennis Roberts) wrote in message 
news:<[EMAIL PROTECTED]>...
> seems like YOU would be the only one to know IF some data value ... is 
> clearly incorrect or in error ... in the case below, i have no idea if 500 
> is a legitimate value or not ...
> 
> if you know that there are clearly wrong values ... eliminate them ...
> 
> if indeed you have millions of values ... what harm can it do?
> 
> by the way ... if you have millions in your sample ... what kind of a 
> sample is it to have SOOOOO many?
> 
> At 12:05 PM 2/18/03 -0800, saisat wrote:
> >Hello all,
> >
> >
> >I have a large quantity of data for which i have the mean and Standard
> >deviation. But in this data, quite of few of the values are really
> >inaccurate. How can I get the "accurate" mean of the above sample
> >data.
> >
> >For example if I have 100, 100, 100, 500 as the sample data
> >the mean is 200. But in my case 500 is obviously an erroneous value
> >for the data. The actual mean should be close to 100. How can I get
> >this value bearing in mind that my sample in quite large (in millions)
> >and i need to somehow "remove" these inaccuracies
> >
> >Thanks
> >Sat
> >.
> >.
> >=================================================================
> >Instructions for joining and leaving this list, remarks about the
> >problem of INAPPROPRIATE MESSAGES, and archives are available at:
> >.                  http://jse.stat.ncsu.edu/                    .
> >=================================================================
> 
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at:
> .                  http://jse.stat.ncsu.edu/                    .
> =================================================================
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to