Databricks Machine Learning Associate Question 257
Single answerYou are building a regression model in Databricks where the target variable represents house prices. To stabilize variance and make the data more normally distributed, you apply a log transformation to the target variable before training your model. After training, you compute the RMSE on the log-transformed scale. What step must you take to interpret the RMSE in terms of the original house price scale?
- A
Exponentiate the RMSE to revert it back to the original scale.
- B
Divide the RMSE by the natural logarithm of 10 to revert it back to the original scale.
- C
Square the RMSE to interpret it in terms of the original scale.
- D
No additional steps are needed since the RMSE already reflects the original scale.
Show answer and explanation
Correct answer: A
Explanation
When the target variable is log-transformed, the RMSE is calculated in the log scale. To interpret the RMSE in terms of the original scale, you must exponentiate it using the same base (e.g., e if using the natural log). This ensures that the RMSE is converted back to a meaningful value in the original target variable's scale, such as house prices in this case.
- A. Correct.
Correct: When the target variable is log-transformed, the RMSE calculated is also in the log scale. To interpret it in the original scale, you need to exponentiate the RMSE.
- B. Incorrect.
Incorrect: Dividing the RMSE by the natural logarithm of 10 is not a valid operation for reverting the log-transformed RMSE to the original scale. This approach would distort the original values.
- C. Incorrect.
Incorrect: Squaring the RMSE is not a valid method for reverting the log-transformed RMSE to the original scale. Squaring is unrelated to the log-transformation process.
- D. Incorrect.
Incorrect: If no additional steps are taken, the RMSE remains in the log-transformed scale and does not reflect the original scale of the target variable.