CMS3508 Assignment Help
Applications of Artificial Intelligence in Cyber Security Assignment help
- Assignment Aims
Aim: to explore and evaluate the use of supervised machine learning to identify cyber attacks through network data.
- Learning Outcomes:
Knowledge and Understanding Outcomes
- Demonstrate knowledge and critical understanding of Artificial Intelligence technology and its application in cyber security.
- Critically assess current practices in the development of Artificial Intelligence in the domain of cyber security.
- Synthesise approaches to the implementation of Artificial Intelligence systems regarding cyber security.
- Critically evaluate and justify the selection and application of Artificial Intelligence technology for a particular application scenario.
- Assessment Brief
You are employed by an organisation developing cyber security monitoring solutions. Given the continuing increase in network attacks, they have recognised an opportunity to start developing a solution to monitor network-based data and detect attacks that are taking place. They are aware that detecting attacks in network data is not possible with traditional procedural programming (e.g., ‘if-else’ statements) because it is not always possible to identify key characteristics that differentiate ‘attack’ and ‘normal’ data. For this reason, they want to consider the use of machine learning technology to learn generalised patterns in the data and discriminate between normal and attack data that might not be easily identifiable by a human. They have chosen to pursue a supervised approach trained on historical datasets. The company is in the early stages of its development, and they want to acquire an understanding of how well supervised learning can categorise the data. You are required to perform a systematic study of different algorithms, compare, and contrast the results, before providing a critical evaluation for them to use
- Establish a strategy as to how you are going to process the data. This could include considering:
- Your approach to trying different algorithms and the use of development environments (e.g., Weka). How many do you plan to try and why?
- Your approach for using training and testing datasets, considering data splitting, cross-validation, etc.
- How you will evaluate and interpret the findings. I.e., what does success look like and what key measurements are you going to use?
- Perform experimentation to acquire key measurements.
- Critically evaluate the suitability of the machine learning application to helping with this analysis task. This should both focus on critiquing the algorithms performance (e.g., time, resource use) and suitability of any findings.