A machine learning and data-oriented quest to screen the degree of long-range order/disorder in polymeric materials
[ X ]
Tarih
2025
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Elsevier
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
In the realm of polymeric materials, the delicate balance between long-range order and disorder dictates crystal properties, influencing their performance in various applications. To unravel this enigma, we embarked on a machine learning (ML) and data-driven quest, compiling 2500 data points from literature. By harnessing the power of Support Vector Machines (SVM) and Radial Basis Functions (RBF), we trained our model to decipher the intricate relationships between molecular descriptors and crystal properties. Introducing a novel pass/fail system, we screened polymers based on their calculated descriptors, revealing that combining multiple descriptors significantly enhances model performance. Identifying 1200 polymers that failed to meet crystallization requirements provides valuable insights for designing materials with tailored structural features. This groundbreaking study pioneers a data-oriented approach to understanding polymeric materials, paving the way for the creation of novel crystals with optimized properties. By uncovering the hidden patterns of order and disorder, we unlock the secrets of polymeric materials, revolutionizing their applications in various fields.
Açıklama
Anahtar Kelimeler
ML, Crystal propensity, Polymer, t-SNE, Support Vector Machine
Kaynak
Materials Today Communications
WoS Q Değeri
Q2
Scopus Q Değeri
Q2
Cilt
43
