Deep Learning Application for Predicting Soil Organic Matter Content by VIS-NIR Spectroscopy

<div>Training and validation curves for model evaluation using the root mean square error (RMSE), coefficient of determination (R<sub>2</sub>), and ratio of performance to deviation (RPD): (a) approach of different models using the full spectrum with 203 variables and (b) approach of different models applied to a subset of 67 selected variables.</div>

Computational Intelligence and Neuroscience

fig6

Figure 6

Figure 6: Deep Learning Application for Predicting Soil Organic Matter Content by VIS-NIR Spectroscopy 