Collaboration (1)

This study will model customer personalities and address common issues in financial data, such as multicollinearity and outliers, by introducing robust logistic regression analysis to predict whether a customer will participate in an event. Due to the imbalance in the predicted target, the commonly used accuracy metric is abandoned in favor of developing a profit-based model evaluation metric to maximize profit as the model selection criterion. Finally, based on principal component selection of variables, the study analyzes and explores the market positioning behind the event and identifies the characteristics of potential customers.