Abstract: Based on rainfall and flood characteristic indicators, this study utilizes runoff and precipitation data from the Zhuanglongwan Station of the Kuye River in the Loess Plateau from 1996 to 2023. It employs principal component analysis and K-means clustering to assess the similarity of rainfall-runoff processes and applies various regression methods to establish quantitative relationships. The study quantitatively analyzes the correlation between three types of floods and their characteristic indicators. The results indicate that for short-duration single-peak floods, multiple linear regression provides the best fit. For long-duration high-peaked floods and small-flow multi-peak floods, stepwise regression yields better fitting results. The conclusions enhance the understanding of rainfall-runoff characteristics in the Yellow River Basin and provide a reference for the efficient utilization of water resources and the prevention and control of droughts and floods.