Mathematical modeling of refractive index based on machine learning (kNN-QSPR) method

In the present work, we created new software in R system, strictly following to the kNN-QSPR approach, used PaDEL and Dragon Descriptor software for generation of initial set of descriptors and developed the program for automatic scaling. Based on this new software, we propose models of quantitative relationships between refractive indices (RI) and polymer structure. An important conclusion of this study is confirmed by the correct interpretation of the constructed models.