Abstract: A machine learning-based method for designing a high-strength high-toughness steel, including: (S1) obtaining data and filling in missing parts to form a data set; (S2) selecting feature data in the data set to form a standard data set; (S3) constructing two machine learning models of the high-strength high-toughness steel; (S4) completing training after the two models are evaluated to be qualified; (S5) finding frontier points, drawing a Pareto front, and distinguishing a known region and a feature space; (S6) in the feature space, setting a step for the feature data, drawing a grid space, and performing multiple training predictions on each grid point by using the models, to obtain predicted Gaussian distributions of two objectives; and (S7) searching for an expected improvement point through an efficient global optimization algorithm, and obtaining design parameter values of corresponding features.