Assessing machine learning models for near-infrared regression by measuring stability towards diffeomorphisms

Near infrared (NIR) spectroscopy is widely used as a tool for non-destructive assessment of fruit quality by applying measured spectra to predict quality parameters such as dry matter and soluble solids content using a suitable regression method. With continued advancements in deep learning, there is potential for improved predictive performance when neural network models are applied instead of partial least-squares regression, but choosing a model remains challenging as performance is sensitive to the model’s architecture. Taking inspiration from work done in image classification, we propose model selection by assessing relative stability to diffeomorphic transformations, providing a complementary approach to standard validation methods. This is particularly useful when labelled validation data is limited. Our empirical results on several NIR regression problems indicate that the proposed approach is comparable to the use of independent validation sets. In addition to the choice of deep learning architecture, we also consider the selection of the number of components in partial least-squares regression to demonstrate the method’s generality. • A novel method for model selection in near-infrared spectroscopy applications is proposed based on stability to diffeomorphic transformations. • The approach can assess model stability on the training data or unlabelled validation data. • Relative stability to diffeomorphisms (Rf) correlates well with test performance across multiple NIR datasets for both deep learning models and PLS regression. • The method can be used to select optimal model architectures and determine the appropriate stopping points during training to prevent overfitting. • Experimental results in mango and kiwifruit datasets demonstrate that the approach is competitive with traditional validation-based model selection methods.