An empirical evaluation of the interpretation methods on Malware analysis

• Main Author: Lee, Andrew Jian Hao
• Other Authors: Liu Yang
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/157254
• Institution: Nanyang Technological University

Malware (malicious software) is a type of software design to damage or abuse any programmable system or network. Most malware do not draw attention to themselves and cannot be seen with the naked eye. Therefore, malware analysis is needed as it is the process of getting to know the behavior and motive of suspicious files or Uniform Resource Locator (URL). Malware analysis can be conducted in 2 manners, static, dynamic, or even both. Static analysis is the testing and evaluation of the internal structure of the application while running it. Dynamic analysis does the total opposite of static analysis where it tests and evaluate on the application during runtime. Throughout the period of my FYP, we will be building up a machine learning model. We will be applying interpretation method of Tensorflow as our source platform for machine learning. To generate our model, we use Keras as training for deep learning models. To evaluate the accuracy of the model, we will be using functional model which allows to build random graphs of layers.