Multivariate regression (MVR) and different artificial neural network (ANN) models developed for optical transparency of conductive polymer nanocomposite films
作者：
Highlights：
• MVR and different ANNs were used to model optical transparency for the first time.
• FFNN model outperformed all other developed network topologies.
• Neural predictive formula was derived using weight and bias values.
• Kernel density mapping showed in which data range much stronger correlations exist.
• Sensitivity analysis results were found to be consistent with correlation matrix.
摘要
•MVR and different ANNs were used to model optical transparency for the first time.•FFNN model outperformed all other developed network topologies.•Neural predictive formula was derived using weight and bias values.•Kernel density mapping showed in which data range much stronger correlations exist.•Sensitivity analysis results were found to be consistent with correlation matrix.
论文关键词：Feed-forward neural network,Generalized regression neural network,Radial basis function neural network,Multivariate regression,Kernel density mapping,Polymer nanocomposites
论文评审过程：Received 16 November 2021, Revised 9 April 2022, Accepted 19 June 2022, Available online 26 June 2022, Version of Record 9 July 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.117937
