Reproduce “Wine Characterisation with Spectral Information and Predictive Artificial Intelligence” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jianping Yao, Son N. Tran, Hieu Nguyen, Samantha Sawyer, Rocco Longo
Researchers at the University of Tasmania developed a framework using UV-Vis spectroscopy and predictive AI to characterize grape juice, successfully predicting sensory attributes with Support Vector Regression and classifying geographical origin with over 91% accuracy using Support Vector Machines.
Key insights
- Support Vector Regression (SVR) emerged as the top-performing model for predicting grape juice astringency, bitterness, and herbaceous characteristics from spectral data.
- Support Vector Machine (SVM) (linear kernel) demonstrated superior performance in classifying grape juice geographical origin (region and vineyard), achieving over 91% accuracy and F1 score.
- The 250-420 nm range of UV-Vis wavelengths was identified as a stable and influential region for both sensory attribute and origin predictions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.