Describe fundamental principles of machine learning on Azure →mediumMultiple ChoiceObjective-mapped

AI-900 Practice Question: Describe fundamental principles of machine learning on Azure

A data scientist trains a regression model to predict daily electricity consumption (in kWh) for a commercial building. The business team needs a metric that heavily penalizes large prediction errors (outliers) more than small errors. Which metric should the data scientist report to best meet this requirement?

⚠ Common exam trap

Many candidates confuse MAE as a robust metric for all error scenarios, but the question explicitly requires heavy penalization of outliers, which only RMSE (or MSE) achieves through squaring errors.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Root Mean Squared Error (RMSE)

Root Mean Squared Error (RMSE) is the correct metric because it squares the residuals before averaging, which disproportionately amplifies the impact of large errors (outliers) compared to small errors. This aligns directly with the business requirement to heavily penalize large prediction errors in the regression model for daily electricity consumption.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
Mean Absolute Error (MAE)
Why it's wrong here
Mean Absolute Error (MAE) averages the absolute differences between predicted and actual values using L1 loss, so every residual contributes linearly and equally to the total error. An outlier with a residual of 50 adds only 50 to MAE, while a typical error of 5 adds 5, meaning the outlier is not weighted any more than its raw magnitude. Because it does not square large residuals, MAE would not penalize outliers more heavily, making it the wrong choice under that objective.
✓
Root Mean Squared Error (RMSE)
Why this is correct
Root Mean Squared Error (RMSE) computes the square root of the mean of squared residuals, directly implementing L2 loss. Squaring each error transforms a 10-unit residual into a contribution of 100 versus 1 for a 1-unit residual, so a single large error can dominate the metric and drive the model to reduce it. This quadratic amplification is exactly the heavier penalty on outliers required, making RMSE the correct metric when large errors are especially undesirable.
✗
R-squared
Why it's wrong here
R-squared (coefficient of determination) reports the proportion of variance in the target variable explained by the model relative to a simple mean baseline, typically as a value from 0 to 1 (or negative for poor fits). Although its residual sum of squares term squares errors, the metric is normalized by total variance, so it measures proportional fit rather than an actionable error magnitude or a training loss that penalizes outliers. Thus it can describe model quality but does not serve as a metric that disproportionately penalizes large prediction errors in the way RMSE does.
✗
Mean Absolute Percentage Error (MAPE)
Why it's wrong here
Mean Absolute Percentage Error (MAPE) expresses each absolute error as a percentage of the corresponding actual value, then averages those percentages. Because the denominator changes per observation, a large absolute error on a high-value target may look small in percentage terms, while a minor absolute error on a tiny target can look enormous, so it does not consistently give larger errors more weight. This relative, scale-dependent behavior means MAPE is not appropriate when the requirement is to penalize large absolute outliers more heavily.

Go deeper

AI-900 Practice Question: Describe fundamental principles of machine learning on Azure

Why each option matters

Option-by-option breakdown

Related to this question