August 2019
Intermediate to advanced
342 pages
9h 35m
English
Once the salient features of the images have been extracted, and the corresponding samples have been classified within their respective classes (spam or ham), it is possible to exploit an SVM to perform model training on these features.
One of the most recent projects on this subject is Image Spam Analysis by Annapurna Sowmya Annadatha (http://scholarworks.sjsu.edu/etd_projects/486), which is characterized by the innovative approach adopted, based on the assumption that the features that characterize a spam image, being computer generated, are different to those associated with an image generated by a camera; and the selective use of SVM, which leads to high accuracy of results compared to a reduced cost in ...
Read now
Unlock full access