<p>Al-Mg-Si alloy has good mechanical properties. However, the traditional trial-and-error experiment is time-consuming and has a large process space. How to efficiently design the heat treatment process is still a huge challenge. In this paper, an alloy process optimization system based on machine learning is proposed to promote the rational design of high-strength and high-toughness Al-Mg-Si alloys. To this end, relevant data of Al-Mg-Si alloys were collected from the literature, and multiple machine learning models were established accordingly. The results show that the optimized regression tree model can well construct the quantitative relationship of material composition–process–performance, and the influence mechanism of composition and process on material properties is analyzed by interpretability. Subsequently, high-throughput screening was performed in virtual samples, and experimental preparation and performance testing were performed within the optimal process interval. The predicted and measured tensile strength of the material in the peak aging stage is 326&#xa0;MPa and 320&#xa0;MPa, and the error is 1.8%. In addition, we also tested the experimental materials by XRD, TEM and SEM and calculated the contribution of each strengthening in the material. The results show that precipitation strengthening is the main strengthening method of the material. Finally, the precipitation strengthening mechanism of the material was revealed by the first-principles calculation.</p>