SCOPE (SEER Clinical Outcome Prediction Expert) calculates the area under the reciever operating characteristic curve (ROC) of each model in one cycle of bianry fusions i.e. SCOPE finds the adjacent risk levels that can be fused to produce a higher area under the ROC.
This program requires the user to provide the observed outcome, original risk factor vector, and the user cut off to call a fitted probability positive (i.e. 1 otherwise 0).
This SCOPE optimizer requires Matlab Stat Toolbox and the following m-scripts downloaded into the same directory:
3. BinaryFusionRecodedFactor=BinaryFusionRecodedFactorColumnforOptimization(tempRisk, i)
%example use: [maxAROC maxAROCFactorVector]=FindMaxAROCBinaryFusionOptimization([1 0 1 1 0]',[2 1 5 3 4]',5,0.5). You could change the input vectors to further test SCOPE.
%files containing the initial risk and AROC and the final risk and AROC after running this command are included here. Notice SCOPE condensed 5 initial risk groups into 2 risk levels.
%Please let me how SCOPE works for you, or if you have questions.
Rex Cheung (2019). SCOPE v.1.0. SEER Clinical Outcome Prediction Expert (https://www.mathworks.com/matlabcentral/fileexchange/37259-scope-v-1-0-seer-clinical-outcome-prediction-expert), MATLAB Central File Exchange. Retrieved .