Generative adversarial networks

Effective Zero-day exploit detection and threat estimation by applying generative adversarial networks on subjective network traffic

In the current age of Information Technology, the scale of technological operations and the usage of technology by common man has been increasing exponentially. However, this also brings an increased level of threats and vulnerabilities in cybersecurity and privacy. In addressing some of the challenges faced in mitigating risk and gaining a competitive advantage over organized cyber threats, I would like to perform research on the “Effective Zero-day exploit detection and threat estimation by applying generative adversarial networks on subjective network traffic”. Cyber threats have always been deterministic until the recent past where the threat and intent have been revealed to for identified cyber incidents. However, ever-growing advancements in cyber technologies and the accessibility to computer networks have only increased the risk of a cyber-attack over the internet.

With crucial aspects of our societies coming from communication to defense and financial systems being digitized, unforeseen threats to computer systems can potentially disturb the fabric of our social structures. Building advanced cyber defense systems equipped with sophisticated technologies such as artificial intelligence can give a competitive advance to defense systems over malicious operations. Having an Adversarial Neural Network in a simulated environment with malicious and defending agents train over the historical network traffic data can be used to find undetected vulnerabilities and exploits. This can be extremely useful in defending against Zero-Day vulnerabilities for a mission-critical computer network in both government private organizations. I have considered one research article addressing key concerns related to day zero attack using applying generative adversarial networks on network traffic

I choose this research topic as it is my area of interest and because of the pressing need for technological advancements in the intersection of cybersecurity and artificial intelligence in defense sectors.  And the courses I have been learning in schools including big data analytics and emerging threats in countermeasures and enterprise risk management are extremely relevant and going to be an asset in my research endeavor. The information governance course which mainly focused on data security is going to help me, deep-dive, and focus on the threats involved in data security aspects.

It is also related to my current vocation where Ab initio data processing and data security integration and configuration developer, where we need to develop various applications enhancements and data analysis related to costumer’s information’s and this research will benefit me understanding emerging threats involved in the workplace related to maintaining the privacy of the user information in terms of applications. I would like to use a qualitative approach for this research methodology. I would like to appreciate an article on Neural Networks and (Dong Jin, July 17, 2020) Auto Encoder Machine The targeted population is people involved in the organization. The finds were done about the previous researches regarding the botnet and new implications involved in the Neural Networks and Auto Encoder Machine (Dong Jin, July 17, 2020) with the concerned emerging threat.

The population targeted is that that engages with the application of information technology devices vulnerable to attacks by viruses whose corresponding antivirus software is not yet available. Quantitative study or research methods will be used to study the topic and evaluate the various trends.

The method will also facilitate expounding or elaboration on the problems associated with the perceptions of different individuals. This will involve the creation of data sets to help in the identification of the Malware. An article related to the topic, Niveditha, Ananthan, Amudha, Sam, and Srinidhi (2020), entails discussing the topic ‘Efficient detection and classification of zero-day malware in big data platforms.’ The study participants involve a group of researchers and respective big data users or rather information technology users that experience or have experienced the zero-day malware attack. The study involved the application of quantitative research approach to reach the respective conclusions. The findings from the research approve that malware samples are gradually increasing with time at a significant pace. This aspect has made identification of the zero-day Malware very difficult.

Bibliography

Dong Jin, J. X. (July 17 2020). Zero-Day Traffic Identification Using One-Dimension Convolutional Neural Networks And Auto Encoder Machine. IEEE Publications , 3.

Niveditha, V. R., Ananthan, T. V., Amudha, S., Sam, D., & Srinidhi, S. (2020). Detect and classify zero-day Malware efficiently in the big data platform. International Journal of Advanced Science and Technology, 29(4s), 1947-1954.