Re-experiment Smart: a Novel Method to Enhance Data-driven Prediction of Mechanical Properties of Epoxy Polymers
This addresses the issue of unreliable data-driven predictions in polymer science, offering a scalable strategy for materials science, though it is incremental as it builds on existing outlier detection and re-experimentation concepts.
The paper tackled the problem of outliers in empirical measurements skewing machine learning predictions for polymer material properties by proposing a method that integrates multi-algorithm outlier detection with selective re-experimentation, resulting in reduced prediction error and improved accuracy with only about 5% of the dataset requiring re-measurement.
Accurate prediction of polymer material properties through data-driven approaches greatly accelerates novel material development by reducing redundant experiments and trial-and-error processes. However, inevitable outliers in empirical measurements can severely skew machine learning results, leading to erroneous prediction models and suboptimal material designs. To address this limitation, we propose a novel approach to enhance dataset quality efficiently by integrating multi-algorithm outlier detection with selective re-experimentation of unreliable outlier cases. To validate the empirical effectiveness of the approach, we systematically construct a new dataset containing 701 measurements of three key mechanical properties: glass transition temperature ($T_g$), tan $δ$ peak, and crosslinking density ($v_{c}$). To demonstrate its general applicability, we report the performance improvements across multiple machine learning models, including Elastic Net, SVR, Random Forest, and TPOT, to predict the three key properties. Our method reliably reduces prediction error (RMSE) and significantly improves accuracy with minimal additional experimental work, requiring only about 5% of the dataset to be re-measured. These findings highlight the importance of data quality enhancement in achieving reliable machine learning applications in polymer science and present a scalable strategy for improving predictive reliability in materials science.