Taylor-feedback deer hunting optimisation algorithm for intrusion detection in cloud using deep maxout network

摘要

Today, cloud computing is a fast emergent computational model and has become popular among users in IT world. It is the distributed computing paradigm that is continually exposed to various threats and attacks of diverse origins. On the other hand, such difficult and distributed model becomes an attractive target for intruders. Identifying the intrusions poses a great challenge for the users and providers of cloud services. Intrusion detection is one of the techniques to protect the cloud operations from severe attacks. Hence, an effective approach is designed using the proposed Taylor-feedback deer hunting optimisation-based deep maxout network (Taylor-FDHO-based deep maxout network) to detect the malicious behaviours in cloud infrastructure. However, the proposed method, named Taylor-FDHO is derived by the integration of Taylor series with feedback artificial tree (FAT) and deer hunting optimisation algorithm (DHOA), respectively. Based on the binary classification step, the process of intrusion detection is accomplished using deep maxout network. However, the proposed approach achieved the maximal accuracy, higher sensitivity, and maximum specificity of 0.9567, 0.9598, and 0.9589 based on the training data.