Analyzing the System Behaviours

Analyzing the System Behaviours

I am happy to announced that one my research articles "Anomaly Detection Techniques Based on Kappa-Pruned Ensembles" has been published in the renowned journal of IEEE Transactions on Reliability.

We proposed an ensemble based machine learning approach that insures the diversities among the fused ensemble classifiers. The approach is tested on two case studies:

(1) We collect and analyze a massive amount of system call traces from the linux based server machines for detecting the anomalous behaviours at the host-level

(2) We also collecting a massive amount of system call traces from the Windows based server machines for detecting the anomalous behaviours at the host-level

The proposed approach significantly reduces the false alarm rates with a higher accuracy

Article is available at: http://ieeexplore.ieee.org/document/8275028


To view or add a comment, sign in

More articles by Shariful Islam, PhD

  • A MapReduce Solution for Profiling Heterogeneous Features

    Since our childhood, we have learned "if we work as a team, a full-day work might be solved just by one minute/second."…

  • Ensemble of Heterogeneous Classifiers

    I want to start with an interesting example. Let say, we are going to detect "Are you handsome? " Labels: 'Yes' (1) or…

  • Smart Online Supermarket

    Few days ago, I visited an online retailer's warehouse in Montreal, Canada. Fortunately, I introduced with an IT expert…

Explore content categories