Improving classifier accuracy #10
Description
I have successfully built a project modeled after the BeanCounter project, but for some reason, every time I try to classify one of the images that I used to train the algorithm, I get the result "nail clubbing" (I'm working with images of fingernail diseases, and nail clubbing is one of the diseases I'm checking for). This Google Drive folder contains all ~110 images I'm using to train, as well as their corresponding classification label. How can I best optimize the app to achieve higher classification rates?