Researchers at Rice University have developed a new machine learning (ML) algorithm that excels at interpreting the “light signatures” (optical spectra) of molecules, materials and disease biomarkers, potentially enabling faster and more precise medical diagnoses and sample analysis.