Wine Quality Prediction with Ensemble Trees: A Unified, Leak-Free Comparative Study

TL;DR

This study benchmarks five ensemble learning methods for wine quality prediction using a rigorous, leak-free workflow on Vinho Verde datasets, providing insights into model performance, feature importance, and computational efficiency.

Contribution

It introduces a comprehensive, reproducible benchmarking pipeline for ensemble models on wine quality data, highlighting the most effective models and feature subsets.

Findings

Gradient Boosting achieves highest accuracy with weighted F1 around 0.69.

Top five features capture most predictive information, reducing dimensionality by 55%.

Random Forest offers the best cost-efficiency for production use.

Abstract

Accurate and reproducible wine-quality assessment is critical for production control yet remains dominated by subjective, labour-intensive tasting panels. We present the first unified benchmark of five ensemble learners (Random Forest, Gradient Boosting, XGBoost, LightGBM, CatBoost) on the canonical Vinho Verde red- and white-wine datasets (1,599 and 4,898 instances, 11 physicochemical attributes). Our leakage-free workflow employs an 80:20 stratified train-test split, five-fold StratifiedGroupKFold within the training set, per-fold standardisation, SMOTE-Tomek resampling, inverse-frequency cost weighting, Optuna hyper-parameter search (120-200 trials per model) and a two-stage feature-selection refit. Final scores on untouched test sets are reported with weighted F1 as the headline metric. Gradient Boosting achieves the highest accuracy (weighted F1 0.693 +/- 0.028 for red and 0.664 +/- 0.016 for white), followed within three percentage points by Random Forest and XGBoost. Limiting each model to its five top-ranked variables lowers dimensionality by 55 percent while reducing weighted F1 by only 2.6 percentage points for red and 3.0 percentage points for white, indicating that alcohol, volatile acidity, sulphates, free SO2 and chlorides capture most predictive signal. Runtime profiling on an EPYC 9K84/H20 node reveals a steep efficiency gradient: Gradient Boosting averages 12 h per five-fold study, XGBoost and LightGBM require 2-3 h, CatBoost 1 h, and Random Forest under 50 min. We therefore recommend Random Forest as the most cost-effective production model, XGBoost and LightGBM as GPU-efficient alternatives, and Gradient Boosting as the accuracy ceiling for offline benchmarking. The fully documented pipeline and metric set provide a reproducible baseline for future work on imbalanced multi-class wine-quality prediction.