NIR-Based Modeling and Analysis of Tobacco Casing Data
Description
This dataset was generated from the experiment on characteristic component identification of Haiyun flavor and construction of near-infrared prediction model in the tobacco flavoring process. The dataset includes raw GC-MS qualitative and quantitative detection data of Haiyun flavor, original near-infrared spectral data of samples, spectral matrix data after different spectral preprocessing, fitting parameter data of various modeling algorithms, comparison data between model predicted values and experimental measured values, as well as statistical results of model accuracy and relative standard deviation. The dataset completely retains the original and derived data throughout spectral preprocessing, algorithm screening, model training and accuracy verification. It can be used to reproduce the identification of flavor characteristic substances and the construction of NIR quantitative model, and support secondary data mining and methodological research in tobacco process optimization, spectral chemometrics modeling and rapid evaluation of flavoring process stability.