Midterm colloquium Sabina Miyuki Ohkawa

Data-Driven Modelling and Analysis of Attention-Working Memory Interplay

Supervisor: dr. Matin Jafarian

Abstract: Both working memory and attention are fundamental concepts in human cognition that are essential for the completion of everyday tasks. Working memory links sensory input with past experiences, while attention is key in filtering this input and protecting memory against distractors. Although they have been studied for decades, many of the neuronal mechanisms which underlie these processes remain unknown. Current understanding has largely been derived either from analysis of experimental data or construction of white-box dynamic models. This work aims to combine these two approaches by applying data-driven modelling to neuroimaging data collected from humans during a working memory task. In particular, new approaches in scientific machine learning, including neural differential equations, show great promise for automated discovery of new scientific models. These models may provide additional insights into how working memory and attention processes interact within the brain.