Provide a comprehensive performance report for both training and testing datasets. Include metrics such as accuracy, AUC, precision, recall, and f1-score for both sets. Standard deviation should be included when reporting the metrics for training dataset. Additionally, present the confusion matrix for the testing dataset.