Prabhjyot's Status Report #12 of 17

[Prabhjyot Singh Sodhi <prabhjyotsingh95 () gmail com>]

Hi list,

This week started with a lot of work which included a lot of reading up and
documentation. Later in the week my mum visited town, so I wasn't able to
get a lot done in the last couple of days.

Achievements this week:
- Read up on avoiding over fitting. Since we are working with very few
prints for training our model. I fear over fitting the model which performs
badly on new, unseen data. To ensure that this doesn't happen, we have used
randomized testing.

- Perform randomized testing using the same randomized tests for all 3
models (liblinear, rfmodel and msrfmodel). This helps compare the models
better. I'll push the test results to nmap-private-dev today.

- Scanned through the existing liblinear prediction implementation to
identify the parts which require changing, when we add the new model for
prediction.

Priorities this week:
- fire a request for comments mail on our current methods and way of going
abou things.
- Convert the opencv makefile to be compatible with autotools that nmap
uses.

Have a productive week folks,
Cheers,
Prabhjyot

_______________________________________________
Sent through the dev mailing list
https://nmap.org/mailman/listinfo/dev
Archived at http://seclists.org/nmap-dev/