Binary Logistic Models: Predicting Enrollments from Application Data

Verified

Added on  2023/01/17

|67
|16852
|55
Report
AI Summary
This report presents an analysis of binary logistic models for predicting student enrollment based on application data. The study employs various regression models, including Logit, to assess the impact of different variables on enrollment probability. The analysis covers multiple subsets of data, considering factors such as program, version, sex, age, nationality, and application characteristics. Key findings include the identification of significant variables influencing enrollment, such as simultaneous applications and numerus fixus. The report evaluates model performance using k-fold cross-validation, Brier scores, and random forest analysis, comparing the performance of different models and highlighting the limitations of the models. The study also discusses the omission of variables due to collinearity and the use of baseline models for comparison, providing insights into the factors that contribute to the success of enrollment prediction models. The report concludes with a comparison of the model's ability to predict enrollment accurately and suggest the best model to choose.
chevron_up_icon
1 out of 67
circle_padding
hide_on_mobile
zoom_out_icon
Loading PDF…
[object Object]