<p>Finite element analysis (FEA) is widely used to analyze the physical and mechanical properties of materials. However, its mathematical precision makes it challenging to handle uncertainties and noise within the data, making it difficult to deduce system inputs from the outputs, which limits its applicability in inverse analysis. Moreover, FEA simulations for complex material problems often require substantial time and computational resources. Machine learning (ML), by learning input–output mappings from training data, can effectively address inverse problems with non-unique solutions. Additionally, it can reduce computational time and costs. Therefore, the integration of ML with FEA can complement their respective strengths and better address real-world challenges in materials engineering. This review systematically explores the applications of ML combined with FEA in areas such as parameter inversion, computational process acceleration and optimization, material property prediction, and material design and optimization. Finally, future directions for the development of the integration of ML and FEA are discussed.</p> Graphical abstract <p></p>