Is there any function similar or better than isequal to solve comparison problem

Hi, I hope somebody can help me eventhough this is kind of algorithm problem using isequal.Let say there is a dataset of 3 columns;col1 stores petal length, col2 stores actual class and col3 stores predicted class.There are 3 classes in this dataset.

PL    actual  predicted
1        2        2
1.1      2        2
1.2      2        2
1.2      3        2
1.3      1        3

Row/Interval 1:3, FP error=0; which is correct.

Row/Interval 4:5, FP error /missclassify produce errors.

FPcost has correct answer for all intervals.

e.g when actual class=1

I expect in interval 4:5, when missclassify as class2 although actual class is 1, FPerror=0

I expect in interval 4:5, when missclassify as class3 although actual class is 1, FPerror=1

My idea is when predicted class is not equal w actual class there is a possibility of missclassification.So, I want to count how many occurences/FPerror of pred_class=2,act_class=1, then pred_class=3,act_class=1

*repeats for each class w different combinations of predicted class and actual class

                [n_act,lablist_act]=classsizes(E(loc:m,2));
                [n_pred,lablist_pred]=classsizes(E(loc:m,3));
                FPerror = 0;
                for k=1:f % for each class
                    disp('FPcost = ');disp(n_act(k));
                    for p = 1 : f-1 % each possibility;if 3class possibility=2
                        for z = loc : m % for each interval
                            tf2 = isequal(E(z,2),E(z,3));
                            if (tf2 == 0) % pred not same w actual 
                                tf3 = isequal(E(z,3),lablist_act(k));
                                tf4 = isequal(E(z,2),k);
                                if ((tf3 == 0) && (tf4 == 1)) 
                                    FPerror = FPerror + 1;
                                end % end tf3 tf4
                            else % pred same w actual
                                FPerror = 0;
                            end %end tf2  
                        end % end z loc m
                        disp('FPerror = ');disp(FPerror);
                        FPerror = 0; % clear value
                    end % end p f-1
                end %for each class

Thank you.