XGB-COF: A machine learning software in Python for predicting the friction coefficient of porous Al-based composites with Extreme Gradient Boosting
Mihail Kolev
Abstract
This paper presents a software called XGB-COF that uses a machine learning algorithm called Extreme Gradient Boosting to predict the coefficient of friction (COF) of porous AlSi10Mg-Al2O3 composites, tested by pin-on-disk method under dry sliding conditions. The software is based on python and uses various packages for data processing, machine learning, and visualization. XGB-COF aims to address the research challenge of developing and enhancing materials with improved wear resistance and low friction for different engineering domains. The software performs hyperparameter tuning using GridSearchCV on the training and validation sets to find the optimal values of the learning rate and the number of estimators. It assesses the model’s performance on the test set using coefficient of determination, squared error, root mean squared error, and mean absolute error. It also generates and stores the plot of the actual vs predicted COF over time, the performance metrics, and the actual and predicted COF data for both test and validation sets. The XGB-COF software is available at https://codeocean.com/capsule/7130568/tree/v1 under an MIT license.