ı want tor plot roc curve for blotch detection. I used ROD detector and ı found a mask image which shows blotch area with white point.
If there is blotch ,the pixel is 1 else 0. So ı compute true positif rate and false positif rate
%% true positive =if there is blotch and blotch detected
%% false positive=if no blotch and blotch detected
%% false negative= if there is blotch and blotch is not detected
%% true negative =if no blotch and blotch is not detected
Rod detector has thres threshold for red green and blue channels( T=[35,135,150];)
what should be the threshold in rank order detector detector for roc curve. I used t =[ 35 135 150; 40 140 155; 45 145 160;
50 150 165; 55 155 170; 60 160 175; 65 165 180
70 170 185; 75 175 190; 80 180 195; 85 185 200] and the result is this and ı think this result is wrong. So what should be the threshold in rank order detector detector for roc curve
hi,
It is the ncredibly useful tool in evaluating and comparing predictive models is the ROCcurve .the (ROC) curve is a two dimensional graph in which the false positive rate is plotted on the X axis and the true positive rate is plotted on the Y axis. ROC detector and found a mask image which shows blotch area with white point , ROC curveis thus the sensitivity as a function of fall-out.
Best Wishes..
hi, thank you for your response C K Gomathy . But ı don't understand your answer?
- Badges
- Science topic
- Similar topics
- Cognitive Science
- Thinking