Abstract: China is one of the countries with frequent rainfall-induced landslide geological disasters. Frequent landslide disasters affect economic development and quality of people's life to some degree. To address this issue, this study proposes a method combining the SHALSTAB model and SVM model, which integrates the advantage of the SHALSTAB model in evaluating the stability of rainfall infiltration on slopes with the advantage of the SVM model in processing nonlinear data. In this study, Daguan County in Yunnan Province has been selected as the research focus to further explore the risk assessment of rainfall-induced landslides. The results show that the evaluation accuracy of the model improves from 82.7% for the single-SHALSTAB model to 94.5% for the SHALSTAB-SVM model, and the accuracy improves by 14.268%. This significant improvement highlights the higher accuracy of the fusion model in accurately analyzing the risk prediction of rainfall-induced landslides. Meanwhile, the satellite image interpretation confirms that the prediction results are consistent with actual landslide conditions, verifying the model's better accuracy in practical applications. This study is not only of great significance for the risk assessment of landslide disasters in Daguan County, but also provides a valuable reference for the prediction and assessment of similar areas or other types of geological disasters.