Research on the Enhancement and Improvement of Logistic Regression Models

The logistic regression model is a linear model widely used for two-category classification problems. This report examines the enhancement and improvement methods of logistic regression models, including regularization techniques, nonlinear extension, mixed effects models, and ensemble learning methods. The regularization technique solves the overfitting problem through L1 and L2 regularization, the nonlinear extension handles the nonlinear relationship through the kernel function, and the mixed effect model solves the data independence and imbalance problems, while the ensemble learning method improves the accuracy and stability of the model. In addition, deep learning methods are also explored to further improve the model performance. The report experimentally verifies the effectiveness of these methods.