Electrochemical profiles of coffee drink samples

Electrochemical Fingerprints and SCA Sensory Profiles of 196 Coffee Samples (Cu/Ni/Au/GC Voltammetry + ML Ground Truth) Authors: Rodion Golovinsky, Ilya Ivanov, Aleksandr S. Aglikov, Timur A. Aliev, Ivan Timofeenko, Alexander S. Novikov, Mikhail F. Zayats, Mariia S. Ashikhmina, Olga Yu. Orlova, Maria S. Masalovich and Ekaterina V. Skorb Description: This dataset enables reproduction of "AI-DRIVEN COFFEE TASTING: PREDICTING TASTE ATTRIBUTES FROM ELECTROCHEMICAL IMPRINTS". Contains 784 cyclic voltammograms (-2V to +2V, 50 mV/s) from 4 unmodified electrodes (Cu, Ni, Au, glassy carbon) measured on 196 ground coffee samples brewed per SCA protocol. Key features:- Raw data: 784 `.edf` files (electrochemical_data.zip) - Processed: 4 CSV tables (5698 current@potential columns, electrochemical_data_csv.zip) - Ground truth: SCA Q Graders scores (coffee_tasting_list_full.xlsx) for 7 attributes (acidity/sweetness/bitterness intensity + quality 6-9 scale) - ML targets: Binary labels - taste (low/middle), quality (good/bad, threshold 7.5) Reproducibility: Achieves F1=0.89 (quality, Au LogisticRegression), 0.87 (acidity, Ni GradientBoosting), 0.72 (sweetness, Cu), 0.63 (bitterness, Au XGBoost). tsfresh features + scikit-learn/CatBoost/XGBoost code: https://github.com/RodionGolovinsky/digital_coffee_tester Non-commercial research only (Creative Commons Attribution Non Commercial 4.0 International). Contact: [e-mail] Files readme.md – detailed description of data structure, file formats, and usage instructions. electrochemical_data_csv.zip – Contains 4 `.csv` files corresponding to the 4 electrode types (`cu`, `au`, `ni`, `gc`) panelists_scores_EN.xlsx – sensory evaluation data (SCA cupping sheets) for ≈200 coffee samples, including tasters’ descriptions and scores for aroma, bouquet, aftertaste, acidity, sweetness, bitterness, body, plus coffee origin, roast, and grind. electrochemical_data.zip – archive with all raw data taken via potentiostat Keywords: coffee tasting, e-tongue, cyclic voltammetry, machine learning, SCA cupping, food authentication, electrochemical fingerprinting