Chartered Financial Analyst (CFA) Practice Exam Level 2

What does the mean squared error (MSE) calculate in model predictions?

1. The average squared difference between actual and predicted values

2. The total number of observations in the dataset

3. The average of the actual values in the dataset

4. The total variance of the dependent variable

The correct answer is: The average squared difference between actual and predicted values

The mean squared error (MSE) is a key measure used to evaluate the accuracy of model predictions by quantifying the average of the squares of the differences between actual values and predicted values. By squaring these differences, MSE emphasizes larger discrepancies, thereby providing a clear metric of how well a model is performing. Calculating MSE involves taking each prediction error (the difference between actual and predicted values), squaring it to eliminate negative values, and then averaging these squared differences across all observations. This results in a single numerical value that represents the extent of the prediction error, allowing for effective comparison between models or tuning of the same model. Other options focus on different statistical concepts that do not relate directly to the assessment of predictive accuracy as MSE does. For instance, one option discusses the number of observations, which is a count and does not inform about prediction errors. Another option looks at the average of actual values, which merely states a central tendency of the dataset rather than evaluating model performance. Lastly, referencing total variance of the dependent variable pertains to the distribution of the data itself rather than its predictive accuracy.