Training Neural Networks by Enhance Grasshopper Optimization Algorithm for Spam Detection System

A significant negative impact of spam e-mail is not limited only to the serious waste resources, time, and efforts, but also increases communications overload cybercrime. Perhaps most damaging aspect email that it has become such a major tool for attacks cross-site scripting, malware infection, phishing, request forgery, etc. Due nature adaptive unsolicited spam, been weakening effect previous discovery techniques. This article proposes new Spam Detection System (SDS) framework, by using series six different variants enhanced Grasshopper Optimization Algorithm (EGOAs), which are investigated combined with Multilayer Perceptron (MLP) purpose advanced detection. In this context, combination MLP EGOAs produces Neural Network (NN) models, referred (EGOAMLPs). The main idea research use train classify emails as non-spam. These models applied SpamBase, SpamAssassin, UK-2011 Webspam benchmark datasets. way, effectiveness our on detecting diverse forms evidenced. results showed proposed model trained achieves higher performance compared other optimization methods in terms accuracy, detection rate, false alarm rate.