Is MSE like Recall?

Question

In one article the classification with using SVM, for each class the MSE(mean square error) is expressed. I'm using confusion matrix. Is that MSE like recall for compare the result of each class?

Answer 1

MSE, which stands for mean_squared_error is a metric that is commonly used in regression problems. It may be used to check the performance of your regressor.

MSE = 1/N * sum(y - y^2)

Recall is a metric, whose value indeed can be deduced from the confusion matrix, which answers the following question : "What proportion/quantity of the actual positives was identified correctly?"

In other words, the mathematical expression is TP / TP + FN, where TP stands for true positives and FN stands for false negatives.

One important distinction is that these metrics are used under different circumstances, the former (MSE) being related to regression problems, while the latter is related to classification problems.