Modelling of Atypical Behaviour Recognition Using Control Dynamics

<p dir="ltr">Imagine a world where your computer doesn’t just respond to your clicks and keystrokes, but understands when you’re confused, frustrated, or even surprisingly delighted. A world where systems could offer help before you even realise you need it, or adapt to your personal learning style based on your genuine emotional state. This is the cutting edge of human-computer interaction, and it’s being made possible by combining the subtle whispers of our bodies with complex mathematical models. This paper dives into how Skin Conductance Response (SCR), interpreted through the lens of Control Dynamics, can help us model and understand those ”atypical” moments in user behaviour, including cyber security measures and principles. An experiment was conducted on thirty (30) participants, where SCR was measured using a mouse pad sensor and webcam, interfaced with eye tracking too, where eye movement data was collected and modelled using a dynamic control system. The results generated by integrating control dynamics into predicting atypical behaviour have demonstrated significant improvements in dynamic control systems over traditional machine learning models, particularly in complex experimental controls and high-stakes environments.</p>