Quantum machine learning for corrosion resistance in stainless steel
Muhamad Akrom, Supriadi Rustad, T. Sutojo, De Rosal Ignatius Moses Setiadi, Hermawan Kresno Dipojono, Ryo Maezono, Moses M. Solomon
Abstract
This study evaluates the efficacy of quantum machine learning (QML) models in predicting stainless steel corrosion behaviour. Using two datasets, the quantum support vector classifier (QSVC) outperformed classical models, achieving accuracies of 95.46 % and 94.80 % for Dataset A and Dataset B, respectively. The QSVC excelled in identifying complex corrosion classes and demonstrated robust performance across diverse environments. This QML approach accurately predicts corrosion without experimental testing, saving significant time and cost. Future research will aim to include more environmental variables and steel types, broadening the model's applicability.