Tell us about Accuracy: advantages, disadvantages, pitfalls.

Accuracy — this is a quality metric of classification that shows the proportion of correctly classified objects out of the total.

Pros:

• Simple and intuitively understandable metric.
• Well-suited when classes are balanced.

Cons and pitfalls:

• Not suitable for imbalanced classes, as high accuracy can be achieved by ignoring rare classes.
• Does not consider the type of errors (false positives or false negatives).

Example: If in a sample 95% of objects belong to one class and 5% to another, a model that always predicts the first class will have an accuracy of 95%, but will not identify the second class.

In such cases, it is better to use metrics that consider class balance, such as F1-score, precision, recall, or ROC-AUC.