Support Vector Machine for Prediction of the Electronic Factors of a Schottky Configuration Interlaid with Pure PVC and Doped by Sm2O3 Nanoparticles

Abstract

This work uses the Support Vector Machine (SVM) to predict the main electronic variables of metal‐semiconductor (MS) and metal‐nanocomposite‐semiconductor (MPS) configurations, i.e., leak current (I0), the height of the potential barrier (ΦB0), ideality coefficient (n), series/shunt resistances (Rs/Rsh), rectification ratio (RR), and surface/interface states density (Nss), along with current conduction/transport mechanisms occurred into them at the reverse/forward biases by analyzing the I–V measurements. The polyvinyl chloride (PVC) and samarium oxide (Sm2O3) nanoparticles are combined to form the two interfacial layers. To analyze the I–V characteristics and train the SVM, the thermionic emission theorem is used. By contrasting the predicted and experimental results, the predictive ability of the SVM approach for predicting the electronic specifications of the fabricated structures and their current conduction/transport processes has been evaluated to investigate the effectiveness of the SVM. There is strong agreement between the experimental data and the SVM predictions of the fundamental electronic characterizations of the MS and MPS structures and the current conduction processes in them at the forward/reverse biases. Additionally, the results demonstrate that the RR value of the MS configuration increases 4 and 53 times if the pure PVC and PVC:Sm2O3 composite interlayers are employed.