949Which solution should the data scientist use to improve the performance of the model?
Apply the Synthetic Minority Oversampling Technique (SMOTE) on the minority class in the training dataset. Retrain the model with the updated training data.
Apply the Synthetic Minority Oversampling Technique (SMOTE) on the majority class in the training dataset. Retrain the model with the updated training data.
Undersample the minority class.
Oversample the majority class.
Explanations:
SMOTE on the minority class (fraudulent transactions) will help balance the dataset and improve the model’s ability to detect fraud. This improves the recall for the minority class (fraud).
SMOTE on the majority class (fair transactions) will increase the number of fair transactions, making the dataset even more imbalanced and worsening the model’s performance for fraud detection.
Undersampling the minority class would remove valuable fraud samples, which would reduce the ability of the model to identify fraud, leading to worse performance.
Oversampling the majority class would worsen the imbalance and reduce the model’s ability to detect the minority class, leading to a decrease in fraud detection performance.
I value that the answer is:
Apply the Synthetic Minority Oversampling Technique (SMOTE) on the minority class in the training dataset. Retrain the model with the updated training data.
I would say the answer is:
Apply the Synthetic Minority Oversampling Technique (SMOTE) on the minority class in the training dataset. Retrain the model with the updated training data.