How can I improve the confidence of my model?

As for improving the confidence, the best way to do this is getting more data. This will, in general, produce terminal nodes based on more examples, thereby tightening the estimates. Additionally, you can improve the confidence level of your model by generating ensembles, engaging in some feature engineering, clustering your readers to find similarities among them before building separate classification or regression models per cluster, or removing anomalous instances in your dataset. You can additionally leverage AI Studio feature selection operator to help you build features which improves overall accuracy and robustness of a model.

You can learn more about Feature Selection Operator here.