Prediction of mango firmness by near infrared spectroscopy tandem with machine learning

The firmness of the mango fruit is one internal physical properties that can show its quality. Unfortunately, non-destructive methods to measure this are not yet available. In current study, we develop a calibration model using near infrared spectroscopy predict (firmness) cultivar Arumanis (Mangifera indica cv. Arumanis) via machine learning. Spectral data were acquired fourier transform near-infrared (FTNIR) benchtop with wavelength range 1000 2500 nm. Multivariate spectra analysis based on learning, including principal component regression (PCR), partial least squares (PLSR), and support vector (SVMR), was utilized compared estimate fresh mangos. results obtained prediction learning by PLSR better than SVMR PCR for firmness. coefficient correlation (rc) validation (rcv), root means square error (RMSE-C) (RMSE-CV), ratio deviation (RPD) 0.941, 0.382 kgf, 0.920, 0.472 2.556, respectively. general satisfactorily indicate technology integrated an appropriate algorithm has optimistic in determining non-destructively.