JDP_V3_N2_RP6 T.R. Jaya Chandra Lekha C. Saravanakumar Journal on Digital Signal Processing 2322–0368 3 2 35 39 NIR, Glucose Detection, SPECTRA Suite, LMA, SVM Diabetes mellitus is a major, and increasing, global problem. The existing method of blood glucose measurement[13] is invasive which requires extraction of blood through a lancing device. This method is painful, potentiality dangerous and expensive to operate. Noninvasive glucose measurement eliminates the painful pricking[11] expensive, risk of infection and damage to finger tissue. Many non-invasive methods for blood glucose monitoring is under study. Optical methods have been developed as the most powerful technique for non-invasive glucose measurement. The NIR spectroscopy method is one of the most promising optical approaches. The spectrum of the blood is obtained from the spectrometer which contains various interfering components. By application of statistical algorithm the interfering substances has to be removed and the peak glucose wavelength has to be determined. The Levenberg-Marquardt algorithm is used to make accurate short-term and long-term blood glucose predictions during the nocturnal period of the daily cycle. April - June 2015 Copyright © 2015 i-manager publications. All rights reserved. i-manager Publications http://www.imanagerpublications.com/Article.aspx?ArticleId=3384