Feature Reduction and Hybrid Machine Learning Model for Email Spam Prediction

Email is a well-known and effectual technique of internet communication and data to share messages. It is employed as an electronic messaging platform to send the messages. However, the increasing significance and higher deployment of emails lead to maximize the number of spam emails. The predictive methods employ distinct phases to predict the email spam in which the data is pre-processed, the attributes are extracted and the data is classified. This work suggests a new mechanism for predicting the email spam. This mechanism employs Principal Component Analysis to mitigate the features. Thereafter, similar type of information is clustered using KMC algorithm. In the end, a voting technique is presented in which K- Nearest Neighbor, LR and Support Vector Machine are implemented. Python is executed to simulate the suggested mechanism. The outcomes are analyzed concerning accuracy, precision and recall.