You could use the accuracy metric from the confusion matrix to determine if the model is overfitting.
The performance can be measured using the percentage of accuracy observed in both training and test datasets to conclude on the presence of overfitting. If the model performs much better on the training set than on the test set, it means that the model is likely overfitting.
This is because your training data is skewed which leads to poor generalization of the model. Ideally, the accuracy on both training and test datasets should be almost equal.
You can go through the following links to read up more about overfitting and the metrics of a confusion matrix :
