Scholarlink

The work presented here investigated the influence of varying weight percentages of a hybrid nano filler composed of Multi-walled carbon nanotubes (MWCNTs) and Polymethyl methacrylate (PMMA) in an epoxy polymer composite. The dielectric properties were analyzed across a 200 MHz to 20 GHz range. The Impact of this hybrid nano-fillers in the epoxy composite on the dielectric properties are investigated. To improve the material design and prediction modeling, machine learning algorithms such as the CatBoost, Histgradient Boosting, LightGBM and Ensemble model are used. In this study, we employed machine learning models to predict the dielectric properties based on frequency and weight percentage. The results indicate that CatBoost and the Ensemble model demonstrated the highest predictive accuracy, achieving the lowest error values and the best fit to actual data. These findings highlight the effectiveness of machine learning in modeling dielectric behavior, offering a reliable approach for predicting material properties with high precision.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning-Assisted Prediction of Dielectric Properties of Epoxy Composites with PMMA/MWCNT Hybrid Nano-Fillers

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sanketsinh Thakor,
  • Prince Jain,
  • Payel Deb,
  • Anand Joshi,
  • Unnati Joshi

摘要

The work presented here investigated the influence of varying weight percentages of a hybrid nano filler composed of Multi-walled carbon nanotubes (MWCNTs) and Polymethyl methacrylate (PMMA) in an epoxy polymer composite. The dielectric properties were analyzed across a 200 MHz to 20 GHz range. The Impact of this hybrid nano-fillers in the epoxy composite on the dielectric properties are investigated. To improve the material design and prediction modeling, machine learning algorithms such as the CatBoost, Histgradient Boosting, LightGBM and Ensemble model are used. In this study, we employed machine learning models to predict the dielectric properties based on frequency and weight percentage. The results indicate that CatBoost and the Ensemble model demonstrated the highest predictive accuracy, achieving the lowest error values and the best fit to actual data. These findings highlight the effectiveness of machine learning in modeling dielectric behavior, offering a reliable approach for predicting material properties with high precision.