Automl

Automl increases MMSE classification balanced accuracy in silico (balanced accuracy 57.9% vs 56.7% and 55.5%).

“AutoML achieved a balanced accuracy of 57.9% in the classification of MMSE, compared to 56.7% and 55.5% for the best benchmark model and ensemble.”

In silicoclassification threshold MMSE >= 29 from 68 cortical thicknesses

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

AAIC 2025 abstract·23 Dec 2025