Question

Does anyone know an algorithms that I can use to calculate the percentage of false positive in a two column list.

Take my situation for instance . I have a clustering vector showing me groups a cluster belongs to and I have the correct label by the side on another column. I know some classifications are wrong from them not mapping to their labels which is most occurring. How can I finding the percentage of false positive for all labels . I am implementing this in R.

Cluster_vector   |    Labels
        1              5     
        3              5
        1              5
        1              5
        6              5

Answer 1

Are you just looking for the proportion of mismatches, like mean(x[,1] != x[,2])?

You can get the confusion matrix by table(x[,1] != x[,2])/nrow(x).