Always getting 'No Tumor' as output for every image.

BrainMRI_GUI.m

Editor's note: Moved the ~700 lines of code to an attachment.

[Image Analyst]

If you have any more questions, then attach your data ('svm_data.mat' and some images) and code to read it in (both the m file and .fig file) with the paperclip icon after you read this:

TUTORIAL: How to ask a question (on Answers) and get a fast answer

[DGM]

Since this is GUIDE, we'd also need the BrainMRI_GUI.fig file that goes with this.

That said, I don't know how you're getting any output. As given, the code won't even run since you've defined two different functions with the same name. (Line 629, 648) Not sure what's going on there.

[Christopher McCausland]

Putting both @Image Analyst and @DGM's very valid points aside for a moment;

If every output is negitive, is the dataset ballanced? i.e. if there are 1000 training observations, but only 10 are true positives (tumors images with matching clinical lable), it would be very easy to train a "always no tumor" output.