Measuring　the　critical　micelle　concentration　(CMC)　of　surfactants　holds　significant　importance　in　comprehending　their　interfacial　properties.　However,　traditional　methods　suffer　from　issues　such　as　lengthy　testing　durations,　low　experimental　accuracy,　and　the　complexity　of　theoretical　calculations.　Herein,　a　method　for　predicting　CMC　is　developed　by　using　machine　learning　(ML)　based　on　the　structural　differentiation　of　surfactants.　A　quantitative　structure-property　relationship　(QSPR)　model　that　can　automatically　classify　and　identify　surfactants　based　on　differences　in　their　head　groups,　was　established　by　collecting　a　diverse　CMC　dataset　of　779　surfactants.　Each　surfactant　molecule　is　quantitatively　chemically　described　using　molecular　descriptors　to　train　5　different　ML　models　by　using　linear　regression　and　tree-based　algorithms.　By　evaluating　model　accuracy,　the　model　was　established　by　automatically　selecting　light　gradient　boosting　machine　(LGBM)　and　gradient　boosting　decision　tree　(GBDT)　as　the　optimal　algorithms　for　ionic　and　nonionic　surfactants,　respectively.　The　overall　prediction　accuracy　of　the　model　achieved　R-2　=　0.944.　Our　model　significantly　outperforms　the　graph　convolutional　neural　network　(GCN)　model　by　comparing　prediction　accuracy　on　the　same　surfactant　data.　Besides,　principal　component　analysis　(PCA)　highlighted　disparities　in　feature　distribution　among　different　types　of　surfactants,　illustrating　the　model＇s　accuracy　and　stability　based　on　structural　variability　and　molecular　descriptors.　This　work　not　only　provides　valuable　insights　into　the　relationship　between　surfactant　molecular　structure　and　CMC　but　also　advances　future　surfactant　design　and　screening.