PDF Malware Detection System based on Machine Learning Algorithm
Pruthvi Priya P M, P Hemavathi · 2022 International Conference on Automation, Computing and Renewable Systems (ICACRS) · 2022
In this digital system, the data is transferred through online in the several form. Safety measures are employed in critical places like healthcare, banks, etc. Nowadays the application of machine learning for solving problems has increased. Adversarial examples are the term used to describe such variations. Early research mostly concentrated on machine learning models for image process after moved to other applications, such as those for malware detection. Finding adversarial examples for ML-based PDF malware detectors is the part of work. Machine learning has apparently delivered extraordinary and, in some cases, human-competitive performance in classification tasks. Hackers can attack and extract the data easily. This proposed work has considered generative adversarial networks (GANs) to build variant PDF malware without any problem that may be identified as benign by using several existing classifiers while preserving the original harmful behavior to address the challenge. Features extraction method, which includes special features derived from malicious PDF files, to quickly produce an evasive variant PDF. The PDF GAN is used for the malware detection in PDF files.