The purpose of this analysis is to find how the results have been biased or impacted for different kinds of ML models from different ML families after being parameter tuning.
- We're used 20 different classifiers from different types of ML Family.
- We're used only Binary Classification.
- Dataset has been found from there.
- After analysis, the results have been published there.
We would like to thank Nedim Srndic and Pavel Laskov for their good work & we're used the same dataset for our malicious pdf analysis that they were used on their project. The research paper can be found here: Practical Evasion of a Learning-Based Classifier: A Case Study. IEEE Symposium on Security and Privacy, 2014 (PDF).