A Survey of Deep Learning techniques for Malware Analysis

  • Molakala Venkata Rakesh Kumar, Anand Kumar S, Ankit Bando, Sachin Raj G.S, Hritik Shah, Shiva Charan Reddy

Abstract

Malware analysis has become an important criterion concerning data security. Due to heavy development in the field of cyber-crime the probability of a component or system being attacked has increased exponentially in the past 5 years. Malware analysis deals with the malware or some Spywares that were attacked on the systems, and with a certain study and proper analysis through the mentioned methodologies, identification and classification are them is made to counter the damage done by it and to prevent any further attacks.

In this paper, we mainly focus on how identification and classification of malware is done through Deep learning paradigms and how it is varied from different approaches. The main advantage of using deep learning is its characteristic of using neural networks of various kinds: CNN, DNN, RNN, multi-layer under various problem statements to make the learning process and to classify them as malware or benign. Traditional methods in malware analysis are implemented by using tools similar to anti-malware. Even the tools require a base algorithm to work on and our goal is to study the algorithms that were suggested previously in this criterion and to suggesting the best one based on the problem domain.

Deep learning in malware analysis can be implemented with various structures like neural networks, random projections, etc. After the structure is implemented feature extraction is done depending upon the dataset of the problem domain and best classifiers must be implemented to the train data models like classification, regression, prediction to increase the accuracy of the model. This survey deals with the analysis of the various machine learning paradigms implemented to overcome malware and to finally propose the best paradigm with the understanding.

Keywords: deep learning, neural networks, CNN, RNN, DNN, data mining.

Published
2020-06-06
How to Cite
Molakala Venkata Rakesh Kumar, Anand Kumar S, Ankit Bando, Sachin Raj G.S, Hritik Shah, Shiva Charan Reddy. (2020). A Survey of Deep Learning techniques for Malware Analysis. International Journal of Advanced Science and Technology, 29(04), 6031 - 6042. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/27206