Machine Learning‐Assisted Optimization of Additive Engineering in FAPbI ₃ ‐Based Perovskite Solar Cells: Achieving High Efficiency and Long‐Term Stability

Additive engineering in perovskite solar cells (PSCs) has been proven to enhance device performance, yet comparing the effects of different additives through experimental methods is still a challenge. Herein, machine learning (ML) is used to quantitatively analyze the impact of additive engineering on performance of PSCs, utilizing a dataset with 778 samples and 39 input features. Key features affecting device performance are identified, revealing that alkali metal additives boost short‐circuit current, alkylamine additives improve open‐circuit voltage, and passivation at A‐site defects is more beneficial than at interstitial sites. Using the results gained from the ML approach, the performance of PSCs improves significantly, achieving an efficiency of 23.50%, with V OC and J SC values of 1.16 V and 25.35 mA cm −2 , respectively, markedly higher than those of the control samples.