CV results of Desicion Tree test mse 0.0000578 and train mse 0

Question

I'm working on a small data set (25 rows, 4 features). I trained the decision tree algorithm and used K-fold cross-validation (cv=3). Then, I obtained r2 0.97 for this reason, I suspected overfitting and looked at the test and train MSE values. I got the test MSE value 0.0000578 and the train mse value 0.0 How should I interpret this fitting do you have over here what do I do? I'm new on this topic :) thank you in advance for your response.

Answer 1

Your dataset is really very small, but everything appears ok. Is the r2 you calculated for your test set? An r2 of 1 would correspond to your model's output being the exact same as the hidden values, and so would an MSE of 0, so your r2 and MSE scores are consistent with one another. The fact that performance is not much worse on the test set than the train set means that you shouldn't be as worried about overfitting, assuming you are splitting your data correctly.