Machine Learning Approach to Polymerization Reaction Engineering: Determining Monomers Reactivity Ratios

TL;DR

This paper presents a machine learning model that predicts monomer reactivity ratios from molecular structures, combining multi-task learning, multi-inputs, and Graph Attention Networks for improved polymerization reaction engineering.

Contribution

The study introduces a novel ML approach integrating GAT and multi-task learning to accurately predict reactivity ratios from monomer structures.

Findings

Model successfully predicts reactivity ratios from molecular structures.

Combines multi-task learning with Graph Attention Networks for enhanced accuracy.

Provides a new tool for polymerization reaction design.

Abstract

Here, we demonstrate how machine learning enables the prediction of comonomers reactivity ratios based on the molecular structure of monomers. We combined multi-task learning, multi-inputs, and Graph Attention Network to build a model capable of predicting reactivity ratios based on the monomers chemical structures.