Dataset on Graphite nanoplatelet enhanced HDPE composites: An ensemble machine learning approach tensile modulus, hardness and toughness
Description
Our research hypothesizes that an ensemble machine learning approach can effectively predict key response variables in polymer composites: tensile modulus, toughness, and hardness. Using a Random Forest regressor, robust to outliers and data distribution, we conducted predictive modeling with hyperparameter tuning via Grid Search CV and Random Search CV. The response variables were categorized into high, medium, and low classes based on quartile distributions, with rules derived using a Decision Tree classifier. The model demonstrated high predictive accuracy, validating our hypothesis. The derived rules provide actionable insights or material optimization. This study's comprehensive dataset and analysis methods lay a solid foundation for future research in polymer composites. These insights can aid in developing high-grade polymer products, optimizing resources, and fostering innovation, significantly impacting the polymer industry.