<p>This study explores the use of advanced optimization algorithms, commonly employed in deep learning, to enhance the performance of solar cells. Specifically, adaptive learning rate methods such as AdaGrad, RMSProp, and Adam are applied to optimize critical parameters of SnS-based thin-film solar cells. Key properties, including the bandgap and thickness of both the absorber and buffer layers, are simultaneously adjusted to maximize cell efficiency. The optimization process utilizes efficiency values calculated by SCAPS, a well-established solar cell simulation tool, as the cost function. This approach offers an efficient and scalable framework for navigating the parameter space of solar cells. The findings highlight the potential of integrating machine learning optimization techniques to advance the design and performance of next-generation photovoltaic devices.</p> Graphical abstract <p></p>