Abstract: Objectives: Accurate susceptibility evaluation results can accurately prevent and control the dangers caused by landslides. Sample optimization is an important method for landslide susceptibility evaluation, which can effectively solve the problem of decision boundary offset generated by unbalanced samples and improve the accuracy of landslide susceptibility evaluation.Methods: Take the southeast area of Wanzhou District of Chongqing as an example, ten influencing factors such as strata, land use and elevation were selected to construct a landslide susceptibility evaluation index system, and the relationship between landslide and index was quantitatively analyzed by applying the frequency ratio method, and on this basis, the deep neural network model (DNN), oversampling-deep neural network model (SMOTE-DNN), hybrid sampling-deep neural network coupling model (OS-DNN), and hybrid sampling-deep neural network-K mean clustering coupling model (OS-DNN-Kmeans) were used to evaluate the landslide susceptibility.Results: The results show that the distance from the road, land use and strata are the main control factors for landslide development in the study area. The accuracy evaluation showed that OS-DNN-Kmeans (95.61%) and OS-DNN (91.16%) could effectively improve the landslide prediction accuracy compared with the models SMOTE-DNN (87.97%) and DNN (81.40%). Conclusions: Sample optimization through mixed sampling and semi-supervised classification can effectively solve the problem of sample imbalance in the study area, and provide new technology support for spatial prediction of landslide disasters.