Sparse and Longitudinal DMDC Classifiers Excel in Simulated HDLSS Data

Sparse Data Maximum Dispersion Classifier (SDMDC) outperformed other sparse models in simulated high-dimensional, low-sample-size data; a longitudinal version (LSDMDC) maintained performance with many predictors, especially under autoregressive correlation. (Simulation only, not…

• Simulation study evaluated classifier extensions on synthetic HDLSS data; no real patient cohort.
• SDMDC showed clear advantage over other sparse classifiers (exact metrics not reported in source).
• LSDMDC handled longitudinal structures best with autoregressive correlation; external clinical applicability remains unproven.

RadPigeon summaries are original and for information only. They are not clinical advice.