I'm trying to find the best C# api/framework for nBayes filtration:
I get some pairs of students' source-code from automatic testing and grading learning system.
- Students can submit code when they think it's correct
- When they submit code it's first version but this submision can be incorrect... then they do some changes and then submit again this is second version (it can be correct or incorrect submission again).
- I identify some specific changes between versions and save them to xml file with information if they lead to correct or incorrect submission. Then I want to apply naive Bayes filtration to predict which changes lead to correct or incorrect answer.
I'm using C#. So far I found nBayes. Is there anything better?