Data‐Driven Design of Polymers with High Melting Point: Chemical Space Generation and Property Prediction

This work is presenting a new method utilizing machine learning to perform the designing of polymers. 10 machine learning models are developed to predict melting points, from which the best‐performing model is chosen for further investigation. An automated process generates 10,000 polymers, and their chemical space is visually explored to uncover underlying patterns. Synthetic accessibility assessment done to select the polymers. Structural characteristics of selected polymers are examined through cluster analysis. This research significantly advances polymer design techniques, providing valuable insights for the development of efficient polymers for various industrial applications.