Hüseyin Camalan: Bayesian Modeling in Neurological Diseases

BCCN Berlin / TU Berlin

 

Abstract

Due to aging populations, the medical field of neurology is facing a substantial challenge with regards to senescence-related diseases, such as dementia. The growing discipline of machine learning (ML) has the potential to transform our understanding of disease mechanisms and to offer practical benefit to populations at risk. This master thesis project constitutes an application of Bayesian modeling to predict disease risk for Alzheimer's disease (AD), the most common type of dementia. The models were validated on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Though the Bayesian models achieved considerable prediction performance, they were slightly outperformed by generic ML models. In addition, some unresolved issues on the model behavior persist."

 

Additional Information

Master Thesis Defense

 

Organized by

Kerstin Ritter