This study further explores clusterwise linear regression (CLR), specifically the expectation maximization (EM) algorithm. The algorithm was quantitatively analyzed through the generated data set. Previous regression methods are integrated and improved to provide better performance in dealing with complex conditions. The model was tested under different independent variables and methodologies, conclusions were drawn from the quantitative results, and optimal solutions under different scenarios were summarized. In addition, the application of the EM algorithm is deduced. The core purpose of this study is to test the reliability and performance of these methods in different situations. By generating diverse data, simulating different environments, and testing various model combinations, we can obtain the optimal regression model combination under different environments. The contribution of this study is to provide an effective and flexible tool for the application of hybrid regression models and provide a reference for future improvements and optimizations. Through this research, we hope to provide new solutions for application scenarios such as market segmentation, medical data analysis, and financial risk management, and provide a theoretical basis for personalized medicine and precision marketing strategies.