-1

I have a maybe naive question of the appropriateness in the use of binary classification. This is a hypothetical example, so forgive me if it is too coarse.

Let’s say I want to train a support vector machine to classify the letter "a" from the letters "b" and "c". The problem is that I don’t have letters "a" to train my SVM model, but I do have a lot of letters "A" which basically mean the same. It is a model like this valid?

My real problem is within the realm of image processing. I have some pixels that I want to classify using texture but I don’t have a "class" for those pixels. I do have ground truth for something very similar (that I know the class) and I have been using the same kind of texture (but from this similar pixels) to train a SVM and do the classification. This SVM works reasonably well (assessed with the visual inspection of the images) with some accuracies up to 90%. The problem I think is that this accuracy is tested with the training of "A" and not "a"

Does this make any sense?

It is something like this valid? or there is anything else that I can code to make this a valid classification? How do you test accuracy in something like this?

PPM
  • 41
  • 4
  • This is a tricky question I would personally just make a dictionary like `{"A":"a","B":"b","C":"c", ....}` which you can check if you found capital A and if so it is related to its lowercase counter part. But to answer your question technically you can. – Flow Oct 13 '22 at 22:24
  • 1
    Not a *programming* question, hence off-topic here; please see the intro and NOTE in https://stackoverflow.com/tags/machine-learning/info – desertnaut Oct 13 '22 at 22:30

1 Answers1

-1

Theoretically a binary classifier can't distinguish between two classes if they are not linearly separable. This is why you get low accuracy. If you want to distinguish between two textures, one approach could be to train a binary classifier for each texture, and then use a voting scheme to make a decision on the label of a new texture.

Sh_gosha
  • 111
  • 2
  • @ Sh_gosha why is 90% low accuracy? Isn't anything above 70% considered "good"? – PPM Oct 14 '22 at 00:40