High-accuracy material classification via reference-free terahertz spectroscopy: Revisiting spectral referencing and feature selection

We investigate how feature selection algorithms can enable accurate, reference-free classification of materials using sparse-frequency terahertz (THz) reflection spectroscopy. This study is based on continuous-wave THz frequency-domain spectroscopy of pellets of molecular solids. The spectra were recorded in a reflection geometry under controlled and ambient humidity conditions. The algorithms were trained and validated on the two datasets, respectively. Three classes of feature selection strategies are evaluated. Namely, the filter-based mRMR (minimum Redundancy Maximum Relevance), the embedded LASSO (Least Absolute Shrinkage and Selection Operator), and the wrapper-based SFS (Sequential Forward Selection) algorithms. Each strategy is assessed using the Linear Logistic Regression, Naïve Bayes, and Support Vector Machine (SVM) classifiers. Our results show that high classification accuracy can be achieved using only a small subset of frequencies. Particularly, when SFS and SVM are applied to non-referenced spectra, 99.5% accuracy was achieved with just 10 features (i.e. 1.5% of the original spectral components). A quantitative comparison of referenced and non-referenced spectra reveals that mirror referencing suppresses reproducible broadband amplitude information that can contribute to classification performance. Furthermore, we show that the SFS-selected features align with the materials’ absorption bands, confirming that the discriminative power arises from genuine spectroscopic contrasts. These findings highlight that reducing spectral dimensionality through data-driven selection eliminates the need for broadband sources and reference measurements, enabling compact, application-specific THz sensors. This approach offers robust material identification in real-world scenarios such as security screening, non-destructive testing, and environmental monitoring. • Feature selection enables accurate classification from raw THz spectra. • Only a few frequencies are sufficient, reducing spectral dimensionality. • Selected features align with material-specific THz absorption bands, confirming physical interpretability. • Eliminates the need for reference spectra in material identification. • Supports development of compact, application-specific THz sensors.