The importance of lateral adsorbate interactions cannot be overstated in describing surface reaction kinetics. To realize the goal of operando computational modeling of catalytic processes, it is crucial to integrate effects of relevant adsorbate coverages and configurations into mean-field kinetic analysis and beyond. Herein, we highlight the recent applications of machine learning (ML) algorithms in the development of adsorbate-adsorbate interaction models, ranging from analytic relationships, to ML-parameterized cluster expansions, and to highly nonlinear deep learning models. We also discuss prospects and challenges in moving the field forward, particularly in the integration of theoretical understanding into ML of lateral adsorbate interactions across the chemistry and materials space.

在描述表面反应动力学时，不能夸大横向吸附物相互作用的重要性。为了实现催化过程的操作计算建模的目标，将相关吸附物覆盖率和配置的影响整合到平均场动力学分析等中至关重要。在这里，我们重点介绍了机器学习 (ML) 算法在吸附质-吸附质相互作用模型开发中的最新应用，范围从解析关系到 ML 参数化集群扩展，以及高度非线性的深度学习模型。我们还讨论了推动该领域发展的前景和挑战，特别是在将理论理解整合到跨化学和材料空间的横向吸附物相互作用的 ML 中。