Sardine drying kinetics dataset and modeling outputs for thin-layer, machine-learning, and CNN-LSTM comparisons

This dataset supports a Journal of Food Engineering manuscript on sardine drying kinetics under solar, hot-air, infrared, instant controlled pressure drop (DIC), and combined DIC-assisted drying treatments. The record includes raw spreadsheets, curated long-format datasets, modeling-ready dry-basis moisture data, thin-layer model outputs, classical machine-learning outputs, CNN-LSTM outputs, statistical comparisons, sensitivity analyses, manuscript figures, reproducibility scripts, and a Journal of Food Engineering-oriented manuscript version. The modeling-ready dataset contains 1787 dry-basis moisture observations from 28 drying curves. Data origins are separated as experimental, derived/curated, or simulated where applicable. Solar simulated curves are provided as derived support data and should not be interpreted as independent physical replicates. Hot-air velocity and curve-specific solar irradiance logs were not available; these limitations are documented in the manuscript and analysis notes. Accepted or in-press article manuscript files are excluded from this deposit until publisher rights and DOI status are fully resolved.