Henan Province has a complex geomorphological type and faces the challenge of frequent landslide disasters. Therefore, efficient and accurate landslide susceptibility mapping (LSM) has great significance to local governments and residents. [Objective] However, in the study of LSM, how to choose the suitable machine learning model for Henan landslide disaster data set to improve the evaluation accuracy still needs further investigation. [Methods] The research focuses on Henan Province, where landslide data is collected and compiled into a landslide disaster database. Using the recursive feature elimination method, 11 factors with the highest relative impact on landslides (slope, elevation, plan curvature, profile curvature, land cover, lithology, soil type, precipitation, road density, river density, fault density) are selected and integrated into a spatial dataset. Then, the models of multilayer perceptron (MLP) neural network , random forest, extreme gradient boosting, and support vector machine were trained, and the performances of the models were evaluated using the receiver operating characteristic curve and the area under the curve (AUC). In the end, we create high precision landslide susceptibility zoning map. [Results] The research results indicate that the MLP model has the strongest adaptability to the landslide disaster dataset in Henan Province, achieving the highest AUC of 0.95. In comparison to SVM, XGBoost, and RF models, the MLP model predicts the smallest proportion of landslide disasters in highly susceptible areas, thus more accurately defining high-risk regions for potential landslide disasters. The predicted extremely high and high susceptibility areas are mainly distributed in the western mountainous and hilly areas of Henan Province, where terrain factors play a dominant role in the development of landslide disasters. [Conclusion] The results can provide a reference for the evaluation of landslide susceptibility with high accuracy in large-scale regions.