Use of Raman spectroscopy to screen diabetes mellitus with machine learning tools: comment

This paper comments on the article Use of Raman spectroscopy to screen diabetes mellitus with machine learning tools by E. Guevara et al. The authors propose an optical method for noninvasive automated screening of type 2 diabetes mellitus. Despite the high performance of the proposed method, results shown by the authors may be ambiguous due to the overestimation of classification models for Raman spectral data analysis. (C) 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement