Re: ML Deployment with WEKA

> I wish to know if it is possible to do ML model deployment in WEKA. If this is possible, kindly point me to helpful texts or online contents.

Yes, you can (in a sense).

But how much work is involved depends very much on what the scenario
is that you want the model to be deployed in, e.g.:
- simple scenario:
  file in, make prediction, file out
- more complex scenario:
  REST request in, make prediction, REST response out

You can load Weka models using your own code and make predictions and
generate the required output:
- Java
  https://waikato.github.io/weka-wiki/using_the_api/
- Python using the python-weka-wrapper3 library
  https://fracpete.github.io/python-weka-wrapper3/

Some gotchas:
- Weka in itself is not thread-safe, you have to ensure that yourself
(ie synchronize all calls to models).
- pww3 does not work properly in multi-process environments used by
Python REST libraries like flask due to its reliance on a JVM running
in the background.

Alternatively, you could also use the ADAMS workflow system
(https://adams.cms.waikato.ac.nz/) as it comes with a large number of
operators for various scenarios (file polling, REST services, etc).
Depending on your requirements, you may only have to write very
minimal code (ie some custom plugins to handle your data) to have it
run as either Linux systemd service or Windows service. Writing ADAMS
flows requires a bit of practice (but there are lots of example flows
available to get you started).
ADAMS derived frameworks have been in use in commercial environments
for processing spectral data (e.g., NIR, MIR, XRF) for quite a number
of years.

Cheers, Peter
-- 
Peter Reutemann
Dept. of Computer Science
University of Waikato, NZ
Mobile +64 22 190 2375
https://www.cs.waikato.ac.nz/~fracpete/
http://www.data-mining.co.nz/
_______________________________________________
Wekalist mailing list -- [email protected]
Send posts to [email protected]
To unsubscribe send an email to [email protected]
To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html