We describe our baseline implementation of linear regression and k-means clustering for predicting a generalized depression score from neural activation data collected from fMRI scans of patients diagnosed with obesity and depression.