Machine Learning Algorithms for Predicting Ceramic Properties in Industrial Design

Abstract

Machine learning models were developed and validated to accurately predict ceramic properties such as mechanical strength and thermal resistance based on material composition and processing conditions, enhancing industrial design applications. By leveraging advanced machine learning techniques, the investigation models industrial ceramic performance, including strength, thermal resistance, and durability, to optimize material selection and manufacturing decisions. Using a combination of experimental data, industrial reports, and public databases, the approach minimizes trial-and-error methods, improving efficiency and reducing costs. The dataset comprises 1,000 to 2,000 data points, representing a diverse range of industrial ceramic materials. Experimental labs test material compositions and conditions, while industrial reports provide real-world performance insights, and public databases validate model accuracy. Among the various predictive models tested, Gradient Boosting Machines (GBMs) demonstrated the highest accuracy and the lowest cross-validation variance, making them the most reliable for real-world applications. The predictive accuracy table revealed minimal discrepancies between estimated and actual values, reinforcing the model’s practicality. The integration of machine learning enhances ceramic material selection, manufacturing parameter optimization, and quality control, making a significant impact on industries such as manufacturing, construction, automotive, and aerospace, where high-performance ceramics are essential. The use of GBMs, Convolutional Neural Networks (CNNs), and Random Forests (RF) in predictive modeling highlights the potential of Artificial Intelligence (AI) in industrial material design, enabling engineers to make more precise, efficient, and cost-effective decisions. Additionally, predictive models contribute to Quality Control (QC) and Process Optimization (PO) by evaluating ceramic materials before manufacturing, reducing defects, and improving product consistency.