연구/산학
PKNU Research 1000
| Lee Seunghun | Designing Metal Thin Film Colors with AI | |||
| 작성자 | 대외홍보센터 | 작성일 | 2025-10-27 |
| 조회수 | 94 | ||
| Lee Seunghun | Designing Metal Thin Film Colors with AI | |||||
 |
대외홍보센터 |  |
2025-10-27 |  |
94 |
Designing Metal Thin Film Colors with AI
A research team led by Professor Lee Seunghun from the Department of Physics at Pukyong National University (President Bae Sang-Hoon) has developed a novel physics-based machine learning model that accurately predicts the color of metal oxide thin films using artificial intelligence (AI).
This research has drawn attention for improving both learning efficiency and prediction accuracy by incorporating the principles of electromagnetics directly into the machine learning algorithm through a strategy known as the “kernel trick.”
The color of metal oxide thin films varies depending on surface microstructure and the degree of oxidation, allowing for the realization of a wide range of colors. However, it has been difficult to quantitatively predict the nonlinear correlations between color and process variables such as oxidation time, temperature, and film thickness.
To overcome these limitations, Professor Lee Seunghun’s team explored a way to incorporate physical principles directly into the internal structure of a machine learning model. They proposed a strategy that improves both learning efficiency and prediction performance by designing the algorithm’s kernel function based on the electromagnetic characteristics of the data.
Professor Lee Seunghun stated, “This study demonstrates that integrating physical understanding into machine learning can enhance both learning efficiency and prediction accuracy, clearly highlighting the importance of physics.” He added, “The concepts and practical examples presented in this research are expected to serve as a foundation for making machine learning more accessible and applicable across various academic disciplines.”
The findings of this study were recently published online in the international journal <Materials Research Letters> (JCR top 7.2% in the field of metallurgical and materials engineering), under the title ‘Optimizing a machine-learning model for color design of metal oxides/metal multilayers with physics-guided kernel trick.’ This research was jointly conducted by Lee Dong-Ik, a master’s student in the Department of Physics at Pukyong National University (the sole first author), and the research team led by Professor Jung Se-Young at Pusan National University.
[https://doi.org/10.1080/21663831.2025.2556752]
