What is the main drawback of using accuracy as an evaluation metric?
Practice Questions
Q1
What is the main drawback of using accuracy as an evaluation metric?
It does not account for class imbalance
It is difficult to calculate
It only applies to binary classification
It does not provide insights into model performance
Questions & Step-by-Step Solutions
What is the main drawback of using accuracy as an evaluation metric?
Step 1: Understand what accuracy means. Accuracy is the percentage of correct predictions made by a model out of all predictions.
Step 2: Recognize what an imbalanced dataset is. An imbalanced dataset has a large difference in the number of examples for different classes (e.g., many more examples of one class than another).
Step 3: Realize that in an imbalanced dataset, a model can achieve high accuracy by mostly predicting the majority class. For example, if 90% of the data belongs to one class, a model could predict that class all the time and still have 90% accuracy.
Step 4: Understand that this high accuracy does not mean the model is performing well. It may be failing to predict the minority class, which could be very important.
Step 5: Conclude that accuracy can be misleading because it does not show how well the model is doing on all classes, especially when the classes are imbalanced.
