Abstract: Based on 200-300 groups data of compositions, processes and mechanical properties, the relationship between the key parameters and tensile property of oxide dispersion strengthened (ODS) alloy was established by machine learning. The results show that the optimum value corresponding to the maximum strength exists in the relationship between the content of Cr, Y2O3, W and Ti and the tensile strength of ODS alloy. The addition of Al has no obvious effect on the increase of tensile strength. Therefore, several optimum compositions of ODS alloy were obtained, and the predicted tensile strength at room temperature is above 1 400 MPa. It can help to promote the optimization of ODS alloy as cladding material for fast reactor application.