MATLAB Draw Roc Curve, and calculate AUC value

Draw ROC curves based on decision values and real labels while calculating AUC values

function AUC = Roc_curve (deci,label_y)%%deci=wx+b, label_y, true label[val,ind] = sort (deci, ' descend '); roc_y = Label_y (i nd); stack_x = cumsum (roc_y = =-1)/sum (roc_y = =-1); stack_y = cumsum (roc_y = = 1)/sum (roc_y = = 1); AUC = SUM (stack_x (2:lengt H (roc_y), 1)-stack_x (1:length (roc_y) -1,1)). *stack_y (2:length (roc_y), 1))        %comment The above lines if using Perfcurve of Statistics Toolbox        %[stack_x,stack_y,thre,auc]=perfcurve (label_y,deci,1);p lot (stack_x,stack_y); Xlabel (' False Positive rate '), Ylabel (' True Positive rate '), title ([' ROC curve of (AUC = ' num2str (AUC) ') '); end

　　

Code from Linzhiren website: https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/#roc_curve_for_binary_svm

function AUC = Plotroc (y,x,params)%plotroc draws the Recevier operating characteristic (ROC) Curve.%%auc = Plotroc (trainin G_label, Training_instance [, Libsvm_options-v Cv_fold])% use cross-validation on training data to get decision values and plot ROC Curve.%%auc = Plotroc (Testing_label, testing_instance, model)% use the given model to predict testing data and obtain decision values% for roc%% example:%% load (' Heart_scale.mat '); % Plotroc (Heart_scale_label, Heart_scale_inst, '-V 5 ');%%[y,x] = Libsvmread (' Heart_scale ');% model = Svmtrain (y,x);% PLO Troc (Y,x,model); rand (' state ', 0); % reset Random seedif nargin < 2help Plotrocreturnelseif isempty (y) | IsEmpty (x) error (' Input data is empty '), ElseIf sum (y = = 1) + sum (y = =-1) ~= Length (y) error (' ROC is only applicable to Bina RY classes with labels 1,-1 '); % Check the Trainig_file is Binaryelseif exist (' params ') && ~ischar (params) model = params; [Predict_label,mse,deci] = Svmpredict (Y,x,model);% the procedure for PREDICTINGAUC = Roc_curve (Deci*model. Label (1), y), ElseIf ~exist (' params ') params = [];end[param,fold] = PROC_ARGV (params); % specify each Parameterif fold <= 1error (' The number of folds must is greater than 1 '); else[deci,label_y] = Get_cv_dec I (Y,x,param,fold); % get the value of decision and label after CROSS-CALIDATIONAUC = Roc_curve (deci,label_y); % plot ROC curveendendendfunction [resu,fold] = PROC_ARGV (params) resu=params;fold=5;if ~isempty (params) && ~        IsEmpty (RegExp (params, '-V ')) [fold_val,fold_start,fold_end] = RegExp (params, '-v\s+\d+ ', ' match ', ' start ', ' End ');            If ~isempty (Fold_val) [Temp1,fold] = Strread ([fold_val{:}], '%s%u ');        ResU ([fold_start:fold_end]) = [];        else error (' Number of CV folds must is specified by "-V cv_fold" '); End Endendfunction [deci,label_y] = Get_cv_deci (prob_y,prob_x,param,nr_fold) l=length (prob_y);d eci = ones (l,1); label_ y = ones (l,1); Rand_ind = Randperm (l); For i=1:nr_fold percent cross Training:folDingtest_ind=rand_ind ([Floor (i-1) *l/nr_fold) +1:floor (i*l/nr_fold)] '); train_ind = [1:l] '; Train_ind (Test_ind) = [];   Model = Svmtrain (prob_y (Train_ind), prob_x (Train_ind,:), param); [Predict_label,mse,subdeci] = Svmpredict (prob_y (Test_ind), prob_x (Test_ind,:), model);d ECI (test_ind) = subdeci.* Model. Label (1); label_y (Test_ind) = Prob_y (test_ind); Endendfunction AUC = Roc_curve (deci,label_y)%%deci=wx+b, label_y, True Label[val,ind] = sort (deci, ' descend '); roc_y = label_y (Ind); stack_x = cumsum (roc_y = =-1)/sum (roc_y = =-1); stack_y = Cumsum (roc_y = = 1)/sum (roc_y = = 1); AUC = SUM (stack_x (2:length (roc_y), 1)-stack_x (1:length (roc_y) -1,1)). *stack_y (2:length ( roc_y), 1))%comment the above lines if using Perfcurve of statistics Toolbox%[stack_x,stack_y,thre,auc]=per Fcurve (label_y,deci,1);p lot (stack_x,stack_y), Xlabel (' False Positive rate '), Ylabel (' True Positive rate '); title ([' ROC curve of (AUC = ' num2str (AUC) ') '); end

Call:

[Y,x] = Libsvmread (' heart_scale.txt ');
Model = Svmtrain (y,x);
Plotroc (Y,x,model);

MATLAB Draw Roc Curve, and calculate AUC value