To investigate the health status of lakes in the Daqing region of Heilongjiang Province and the differences in their dominant influencing factors, the K-means clustering machine learning classification algorithm was employed to group 60 typical lakes in the Daqing region based on their attribute characteristics. This approach explored the common typical features within each group and the differences between groups. An evaluation indicator system was established. Leveraging the high-dimensional nonlinear mapping capability of the random forest algorithm, lake health evaluation models were constructed for each cluster based on the grouping results. These models assessed lake health within each cluster and identified the dominant influencing factors for lake health within each group. The research findings indicate:The typical lakes in the study area can be classified into five clusters based on their attribute characteristics. This grouping can be used to enhance the accuracy of the random forest model for health assessment. The typical features of each cluster are:high phytoplankton diversity index, low vegetation coverage, high water self-purification capacity, high fish stock index, and excellent water quality category. The lake health assessment model based on the random forest algorithm can accurately evaluate the health status of typical lakes in the Daqing region while avoiding the shortcomings of traditional evaluation methods, which are prone to significant subjectivity. The constructed evaluation indicator system comprehensively considers the factors influencing lake health. The newly added evaluation indicators, such as lake area shrinkage ratio