I am using ROC method to evaluate the classification stage my question is the area under curve represent the accuracy as well as how can I use ROC method to evaluate the segmentation if I have standard segmentation dataset
Regards
I will presume that you mean the area under the ROC curve (AUROC).
The first thing to do is interpret the AUROC. You will read that the AUROC is the probability that any randomly positive example will have a higher score than a randomly selected negative sample. So with this interpretation, AUROC is related to ranking, and this doesn't have an obvious link to accuracy yet.
It can be shown that the AUROC is related to accuracy in the following manner:
[math]\mathbb{E}(acc) = \frac{AUROC}{2} + \frac{1}{4}[/math]
for even class ratios, and in this manner we can see the relationship between the two - they aren't directly equivalent, but there is a relationship.
For more information about ROC/AUROC analyses, I recommend reading "An introduction to ROC analysis" by Tom Fawcett, see link below.
http://people.inf.elte.hu/kiss/13dwhdm/roc.pdf
Please read the book by Nathalie Japkowicz and Shah: Evaluating Learning Algorithms (Cambridge). You can find there plenty of algorithms and methods used for classification validation.
Cheers
- Badges
- Science topic
- Similar topics
- Cell Biology
- Methods