This interactive tool helps you explore how model complexity (Decision Tree depth), training sample size, and data dimensionality (feature set) shape performance on UCI Irvine – Predict Students’ Dropout and Academic Success dataset. Adjust the controls and see how overfitting/underfitting emerge. Extensive experimentation will help you build an intuition for diagnosing and fixing models when accuracy is less than desirable.
The following Towards Data Science article discusses the insights I took away from experimenting with this tool.
Two independent instances are provided side by side to facilitate comparisons between different configurations.