The role of machine learning in perovskite solar cell research

Over the last few years there has been an increasing number of papers using machine learning (ML) as a tool to aid research directed towards perovskite solar cells. This review provides an overview of this recent development with a focus on the type of questions and narratives being explored, which type of data that may be useful, and how to extract relevant features for training ML models. Key areas being discussed include how to make better sense of experimental data, how to automate and speed up experimentation and data analysis, and how to accelerate theoretical screening for interesting new materials. The primary target group for this review is everyone interested in using ML for material science, regardless of prior ML experience, and who either would like to have an overview of how ML have been used in perovskite research or find inspiration for designing new ML based projects.& COPY; 2023 Elsevier B.V. All rights reserved.

Automated summary

In recent years, there has been a growing number of research papers utilizing machine learning in perovskite solar cell studies. This review provides an overview of this development, focusing on the questions and narratives explored, types of useful data, and the extraction of relevant features for ML models. Key areas discussed include better understanding of experimental data, automation and acceleration of experimentation and data analysis, and faster theoretical screening for new materials. The target audience for this review is anyone interested in using ML for material science, regardless of prior ML experience, who would like to gain an overview of ML applications in perovskite research or seek inspiration for new ML-based projects.