Predicting Reactivity with Machine Learning
METADATA ONLY
Loading...
Author / Producer
Date
2025
Publication Type
Book Chapter
ETH Bibliography
yes
Citations
Scopus:
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
The integration of machine learning into chemistry has revolutionized the field, enabling experimental and digital chemists to predict and elucidate reactions more efficiently. This chapter focuses on the recent applications of data-driven models to predict various aspects of chemical reactivity, such as yields, activation energies, rate constants, selectivities, and catalytic turnover frequencies. We will provide an overview of diverse predictive methods, from traditional approaches like quantitative structure-activity relationships to modern deep learning techniques, as well as strategies specific to catalysis, such as volcano plots. Despite significant advancements, prominent obstacles remain, among them data quality and model generalizability. We will conclude this review by highlighting these challenges and possible ways to overcome them.
Permanent link
Publication status
published
Book title
Artificial Intelligence in Catalysis: Experimental and Computational Methodologies
Journal / series
Volume
Pages / Article No.
157 - 194
Publisher
Wiley-VCH
Event
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Machine learning; Reactivity; Prediction; Modeling; Homogeneous catalysis; Catalyst design
Organisational unit
09781 - Jorner, Kjell / Jorner, Kjell