Machine Learning for Cloud Security: A Systematic Review

The popularity and usage of Cloud computing is increasing rapidly. Several companies are investing in this field either for their own use or to provide it as a service others. One the results development emergence various security problems both industry consumer. ways secure by using Machine Learning (ML). ML techniques have been used prevent detect attacks gaps on Cloud. In paper, we Systematic Literature Review (SLR) methodologies techniques. We analyzed 63 relevant studies SLR categorized into three main research areas: (i) different types threats, (ii) used, (iii) performance outcomes. defined 11 areas. Moreover, distributed denial-of-service (DDoS) data privacy most common areas, with 16% level 14%respectively. On other hand, found 30 some hybrid others standalone. popular SVM standalone models. Furthermore, 60% papers compared models prove efficiency proposed model. 13 evaluation metrics were enumerated. applied metric true positive rate least training time. Lastly, from 20 datasets found, KDD CUP'99 among studies.