One of my reviewers put a comment to my paper as "Show pictures of a real experimental test.". However, my paper describes a machine learning technique in electrical engineering. I don't have any real experiments and I don't claim anything in the paper about that. I can not provide any result regarding the real experiment. The other comments are good and I worry about this comment. Could you help me to provide a response for that?
There are two obvious possibilities here.
The review says only irrelevant things. In that case, simply respond to the editor: "I was unable to determine the relationship between the comments of Reviewer 1 and the submitted manuscript. Clarification would be appreciated." The editor will conclude the referee did not read the paper and the review will be ignored.
The review is unclear due to sloppy writing or limited English skills. In that case, I would interpret the comment as "Add a graph displaying evidence that the paper is correct." The response would be "We have added a graph displaying y as a function of x, which supports our conclusion that ..."
Also, when you are a peer reviewer, try not to give vague feedback.
If you don't intend to do any real experiments any time soon, the best response is for you to be honest in your reply. Explain again that your work consists of a theoretical technique and that (if it's the case) it will be validated experimentally in future works. Maybe support it with a new graph (as Anonymous Physicist suggested), maybe change some paragraphs/sentences to make your point clearer.
Sometimes, reviewers don't really pay attention to details. It has happened to me to be questioned by reviewers about points that I had specifically explained in the text, as if the reviewer hadn't read that part of the article. So maybe he didn't understand your work is theoretical, or maybe there's a sentence that created some confusion.
By being honest, you'll either have your paper accepted (since you mentioned that the other comments are good) or rejected if the editor considers that experiments are necessary. If it is rejected, it is not necessarily a bad thing: most of the times, a rejection is useful for you to improve your paper and resubmit, and you can even get accepted in a better journal/conference with the same work!