Automated and Unbiased Machine Learning Predictions of Alzheimer's Risk Factors in an Ethnically Diverse Dataset using Cortical Thicknesses

In the challenging task of predicting risk factors exclusively from cortical thickness values, AutoML demonstrated its ability to adapt to each task. Despite identical input data, AutoML selected unique pipelines for 5 out of 7 tasks, highlighting the lack of a universal model and showing the potential value of AutoML in Alzheimer's research.