Today’s grand challenges of the polymer community are to improve polymer technologies and deployment faster and more cost efficient than before. Design and discovery strategies of the nascent field of polymer informatics are instrumental to overcome this challenge. Big data and classical modeling are only beginning to impact polymer research and industry but have shown that they act as the shining beacon for many future discoveries.

This symposium focuses on innovative methods for polymer data generation, data curation, knowledge discovery, property predictions, and intelligent design as well as synthesis for pure polymers, polymer blends, and formulations. The purpose of this symposium is to bring together materials scientists/engineers, chemists, physicists, and computer scientists across academia and industry to discuss current and future informatics efforts that advance polymer research. The scope of the discussion includes state-of-the-art methods, recent progress, and cutting edge approaches in classical and data-driven informatics. Potential classical methods include but are not limited to numerical approaches for differential equations, monte carlo simulations, molecular dynamics, and density functional theory. Data-driven approaches include all machine learning methods ranging from more traditional methods such as kernel ridgid, gaussian processes, and random forests to sophisticated neural networks architectures such as variational autoencoder, transformers, physics- or chemistry-informed models, or reinforcement learning.

polymer