Abstract
Machine learning (ML) algorithms based on mathematical and statistical methods are being extensively used for mimicking human decisions in the industry. The current analysis employs and compares the use of four different machine learning algorithms (decision tree, support vector regression (SVR), gaussian process regression (GPR), and ensemble method) for the prediction of the tensile strength of carbon and glass fiber reinforced thermoplastic composites. Such composites find a wider range of applications in the aviation, transport, and sports sector for their extended benefits. The ML model was developed using the MATLAB tool that predicted the tensile strength of the polymer matrix composites based on the five input features. The models were compared based on the performance metrics of mean absolute error (MAE), mean squared error (MSE), root mean square error (RMSE), and coefficient of determination (R2). GPR was the most reliable algorithm as it attained the highest R2 value of 0.98 and the least error values among all other algorithms.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call