N Sukumar - Prediction of polymer properties using infinite chain descriptors (ICD) and machine learning: Toward optimized dielectric polymeric materials

Document created by N Sukumar on Aug 8, 2016
Version 1Show Document
  • View in full screen mode

  Publication Details (including relevant citation   information):

  J. Polym. Sci., Part B: Polym. Phys. 2016. DOI:  


  To facilitate the development of new polymeric materials, we   report the development of new heuristic models to predict the   dielectric constant, band gap, dielectric loss tangent, and glass   transition temperatures for organic polymers. A new set of   features called infinite chain descriptors (ICDs) was designed   and developed especially to characterize organic polymers,   utilizing methods with minimal dependence on predefined fragment   libraries. Machine learning models were built for the   aforementioned properties incorporating best practices in the   field such as objective feature selection, cross-validation and   external test sets. All models produced in this study showed good   performance in prediction. A web tool has been developed and has   been made available that supports the input of novel structures.

  Address (URL): http://onlinelibrary.wiley.com/doi/10.1002/polb.24117/full