XII International Symposium ''Optics and Biophotonics'' (Saratov Fall Meeting 2024), Saratov, September 23-27, 2024
Theoretical and Mathematical Physics

Application of machine learning methods to predict the optical absorption coefficient of composite ceramics based on hydroxyapatite

A. E. Rezvanova^a, B. S. Kudryashov^a, A. N. Ponomarev<sup>ab

a</sup> Institute of Strength Physics and Materials Science, Siberian Branch of the Russian Academy of Sciences, Tomsk, Russia
^b Tomsk State University of Control Systems and Radioelectronics, Tomsk, Russia

Abstract: Models for predicting the optical absorption coefficient of hydroxyapatite-based ceramics and composites with additives of 0.1 and 0.5 wt.% multi-walled carbon nanotubes additives in the terahertz radiation frequency range from 0.2 to 1.4 THz were constructed based on experimental data using machine learning methods. The lowest value of the mean absolute error was shown by modeling using methods of adaptive boosting (0.951%) and neural networks (0.049%). The results of numerical simulation confirm that the use of machine learning methods makes it possible to predict the absorption coefficient with high accuracy for ceramic materials with carbon nanotube additives in the range from 0 to 0.5 wt.% concentrations. The obtained results make it possible to optimize the composition of hydroxyapatite-based ceramics to control their optical characteristics.

Keywords: prediction, regression analysis, machine learning, neural networks, hydroxyapatite, multi-walled carbon nanotubes, absorption coefficient.

Received: 15.01.2025
Revised: 15.01.2025
Accepted: 15.01.2025

DOI: 10.61011/JTF.2025.05.60277.3-25

Bibliographic databases:

•