Databricks Machine Learning Associate Question 254
Single answerA data scientist trains a regression model to predict housing prices but applies a logarithmic transformation to the target variable (house prices) before training. After evaluating the model, they calculate the RMSE in the log-transformed space. Why might it be necessary to exponentiate the RMSE in this scenario?
- A
To convert the RMSE back to the original scale of the target variable (house prices).
- B
To make the RMSE easier to interpret in terms of percentages.
- C
To ensure that the RMSE calculation is unbiased and accurate in the log-transformed space.
- D
To avoid negative values in the RMSE calculation.
Show answer and explanation
Correct answer: A
Explanation
When the target variable is log-transformed during model training, the RMSE is calculated in the log-transformed space. To interpret the RMSE in terms of the original scale of the target variable (e.g., housing prices), it must be exponentiated. This ensures that the error metric is meaningful and directly comparable to the original data.
- A. Correct.
Exponentiating the RMSE transforms it from the log-transformed space back to the original scale of the target variable, making it interpretable in terms of the original housing prices.
- B. Incorrect.
Exponentiating does not convert the RMSE into percentages. It simply reverts it to the original scale of the target variable.
- C. Incorrect.
The RMSE calculation itself is already unbiased and valid in the log-transformed space. Exponentiating is not required for accuracy but for interpretability.
- D. Incorrect.
The RMSE is a non-negative value by definition, so exponentiating is unrelated to avoiding negative values.