[mlpack] Is scanning business transactions for fraud an appropriate use of MLpack?

[Rick Hedin]

Hi.  Could you give me a reading on whether MLpack is an appropriate tool
for what I want to do?  Too often, you start down a path, and after a few
weeks you realize "Oh.  I shouldn't be doing this."

At a previous company, I made an AI system using CRM114.  CRM114 did
Bayesian analysis on data.  We would feed it websites, and it would
classify the websites as being about entertainment, sports, news, music, or
whatever.  There were about 20 categories.  We also had a training
interface, where our Product Owner would feed in websites and tell CRM
"That's a news website," or "That's an entertainment website."  We used the
categorization to present sports websites to a customer who had expressed
an interest in sports, for example.

At my current company, we exchange messages about business transactions
with a lot of other companies.  A certain fraction of the transactions are
fraudulent.  Someone claims to have provided services to a customer and
wants to be paid, but there is no customer with that name, for example.  I
would like to put an AI process on the message stream, transparent to other
uses of the message stream.  When one of our operators marks a transaction
as "possibly fraudulent," that would be a data item for the AI process.
When they later mark it "definitely fraudulent" or "definitely not
fraudulent," those are also data items for the AI.  Eventually, the AI
would be able to add additional tags in the record "AI suspects this
transaction is fraudulent" or "AI suspects this transaction is not
fraudulent," along with "AI confidence is xxx%."

The nice thing about this setup is nobody has to spend hours training it.
The data stream provides both data, and judgement on the data.

So, is this a good application for MLpack?  Or is it more intended for
other purposes, and a different software suite is more appropriate?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://knife.lugatgt.org/pipermail/mlpack/attachments/20181106/255e8429/attachment.html>