Dementia is a diverse set of diseases characterized by a progressive cognitive decline, behaviour change, loss of functional and social ability and high five-year mortality rates.
Temporal changes in these observable features can be thought of as disease trajectories or dementia phenotypes. Clearly classifying these diverse and informative trajectories is central to research on the causes of dementia and to provide effective care for those with dementia. This project aims to develop and apply both model-based and algorithm-based statistical and machine learning methods to longitudinal trajectory clustering of multiple repeatedly measured features of cognitive decline.
The student will conduct longitudinal trajectory analyses under the guidance of the supervisor and project PIs to identify dementia trajectory clusters with multiple longitudinal features using various statistical approaches, including group-based trajectory modeling and other finite mixture models.