SDAV 4: Machine Learning

In this episode we conclude the course, and highlight just some of the future research directions where security data analytics and visualisation will continue to play an important role in our modern technological society.

In this episode we explore how file data can be visualised and analysed, as well as performing more sophisticated analysis on rich image and video content to detect similar people and scene changes.

In this episode we explore text analytics, and how this can be used in security analysis, from models of next-word prediction based on word occurrences, document and spam detection, recommendation systems, and recent advances of GPT language models.

In this episode we explore the meaning of Visual Analytics, and begin to consider human-machine collaborative teaming and interactivity methods for deeper examination of data.

In this episode we discuss visualisation techniques, their strengthens and weaknesses, and how these can be used in the cyber security domain.

In this episode we start to explore what Machine Learning is, and how different methods can be used.

In this episode we start to explore what features about data we made need to make security decisions, and how we collate statistics to inform this. We also look at why statistics alone may not tell the full story about our data, and why correlation may not always be reliable either.

In this episode we talk about the pipeline process of analytics, and we examine other models of data science workflows from academic literature.

In this episode we give an introduction to Security Data Analytics and Visualisation, where we explore why data science is a crucial part of modern cyber security, how visualisation can be used to both explore and explain our data findings, and introducing how artificial intelligence and machine learning are being used in cyber security data analytics.