A Review on Machine Learning Techniques for Image Based Spam Emails Detection

Sending and receiving e-mails have continued to take the lead being the easiest and fastest way of e-communication despite the presence of other forms of e-communication such as social networking. The rise in online transactions through email has globally contributed to the increasing rate of spam emails relatively which has been a major problem in the field of computing. In this note, there are many machine learning techniques available for detecting these unwanted spams. In spite of the significant progress made in the figures of literature reviewed, there is no machine learning method that has achieve 100% accuracy. Each algorithm only utilizes limited features and properties for classification. Therefore, identifying the best algorithm is an important task as their strengths need to be weighed against their limitations. In this paper we explored different machine learning techniques relevant to the spam detection and discussed the contributions provided by researchers for controlling the spamming problem using machine learning classifiers by conducting a comparative study of the selected machine learning algorithms such as: Naive Bayes, Clustering techniques, Random Forest, Decision Tree and Support Vector Machine (SVM).