Ransomware Detection by Machine Learning in Cloud with Source Code and Documentation

Project Discription

Ransomwarre in cloud makes serious losses for service provider and user both. So detection of such attacks need to be done at each session in the network. This paper has proposed a genetic and neural network hybrid model for ransomware detection. It was obtained that proposed model has reduced the input training dataset by use of math flame optimization genetic algorithm and further paper has trained the neural network from selected feature set. Experiment was done on real ransomware dataset different testing sessions. Result shows that model has increases the precision by 13.9% as compared to existing model. It was obtained that proposed model has increases the performance by 11.5% as compared to other existing models. Moth flame based feature optimization has increases the performance of the work. Learning of neural network with less and effective number of features increases the work performance.

Project Sample Image