Title: High-quality liquor prediction through machine learning and PCA advancements

Authors: Rupa Rani; Aman Sanger

Addresses: Department of Computer Science and Engineering, Ajay Kumar Garg Engineering College, Ghaziabad, U.P., India ' AventIQ Private Limited, Mohan Cooperative Industrial Estate, New Delhi, India

Abstract: Machine learning is increasingly applied in diverse industries, including healthcare, astronomy, and hospitality. In the growing liquor industry, predicting high-quality liquor using machine learning can significantly enhance production efficiency and reduce costs. This study proposes a model that combines random forest (RF) with principal component analysis (PCA) to predict liquor quality based on chemical properties, grape growth conditions, and production processes. The model achieves a high accuracy of 91.68%, outperforming traditional methods by effectively capturing non-linear feature interactions. PCA is used to reduce dimensionality and balance the dataset, making the prediction process more efficient. The proposed RFPCA model provides deeper insights into the variables influencing liquor quality and demonstrates better performance compared to previous studies. This approach not only supports quality assessment but also helps liquor manufacturers maintain consistency and control over production. Additionally, demographic analysis assists in identifying ingredients that affect the quality, ensuring improved consumer satisfaction.

Keywords: machine learning; high-quality liquor; random forest principal component analysis; RFPCA; decision tree algorithm; SVM; logistic regression.

DOI: 10.1504/IJQET.2025.151226

International Journal of Quality Engineering and Technology, 2025 Vol.11 No.1, pp.47 - 70

Accepted: 28 Jul 2025
Published online: 19 Jan 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article