We used data from 202 participants from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and 77 participants from the placebo arm of the phase 3 trial of semagacestat for AD (LFAN) with diagnosis of mild-to-moderate AD at baseline. K-Nearest Neighborhood (kNN) models were trained using ADNI data to classify participants to 1) cognitive decliners, change of >0 from baseline in the Alzheimer’s Disease Assessment Scale cognitive subscale (ADAS-cog) score; and 2) Stable cognition, change of ≤0 from baseline ADAS-cog score, after 12 months of follow-up. Trained models were applied to data from LFAN trial (validation sample) to predict disease progression in 12 months. Feature-set in all models included demographics, Apo4 status, neurocognitive tests, and volumetric MRI.