Support Vector Machine and Feature Selection Based Optimization Framework for Big Data Security

As the Internet is rising at a rapid pace, cyber attacks have also increased as a result. There is a rapid rise in the form and rate of incidence of these attacks. There are several conventional security solutions that exist, but in the case of big data, these solutions do not work well. Instead of conventional Processes, shielding big data from attacks requires a different approach. The spark tool is used in this article and it has several advanced features, such as parallel data processing. Some inbuilt machine learning algorithms are included in its library. The dataset used is NSL KDDCUP, the scale of which is MBs. This dataset is ideally suited as a test case for big data. An approach is suggested for intrusion detection, i.e., SVM classifier is used and a new form of approach called Pareto fronts multi objective genetic algorithm is used to pick features