Title: A regularized model for spectroscopic data
Authors: Chin Gi Soh - Nanyang Technological University (National Institute of Education) (Singapore) [presenting]
Ying Zhu - National Institute of Education, Nanyang Technological University (Singapore)
Abstract: High-dimensional spectroscopic data is informative, and has applications in many fields such as biomedical sciences and food science. The fitting of regression models for the purposes of prediction is known to be a challenging task due to the high dimension of the datasets, as well as the high correlation between wavenumbers in the data. One method that has gained interest in recent years is the use of regularization to overcome these challenges. We present a regularized model for spectroscopic data. The penalty functions used are designed to capture the underlying group structure in the spectroscopic data, as well as to give rise to an interpretable model. Depending on the penalty functions used in the regularized model, different computational challenges may arise. We will discuss some algorithms that are of interest in solving for such regularized model coefficients, as well as the advantages and disadvantages of these algorithms. An example of the application of the model to Fourier-transform infrared spectroscopic data for the prediction of olive oil purity will be presented.