MLNIRdata: 208 near-infrared spectroscopic spectra and densities of hydrocarbon mixtures for chemometrics, data science, machine learning or signal processing

This note describes the content of the MLNIRdata dataset. It has already been used in chemometrics for property prediction of chemical mixtures with tools like "Partial Least Squares" (PLS) or sparse PLS. Its publication as "open data" is meant for further analyses and benchmarks in chemometrics, data science, machine learning, signal processing or artificial intelligence applications (prediction, regression, clustering, training, etc.). Its formats (including "csv" files) can be imported into standard data processing frameworks (Matlab, Python, Julia, R). It is available at https://doi.org/10.5281/zenodo.16781223.