In my discussion over my question on Math SE, I explained to a user, how I think AI works, I wrote that with the sigmoid(logistic) function, features of a data set are identified, many such iterations provide learning.

Is my understanding of how this works correct?

  • $\begingroup$ Artificial Intelligence as a field is not only neural networks, you are loosely describing a neural network, not describing AI in itself. Please do not call neural networks as "AI". $\endgroup$ – Dr. Snoopy May 16 '20 at 9:44
  • $\begingroup$ What other purpose does neural networks other than to create AI? $\endgroup$ – noviceFedora May 16 '20 at 16:14
  • $\begingroup$ Neural networks can do curve fitting, would you call that AI? Please do not be naive, AI is much bigger than just neural networks. If you look at Alpha Go, you can easily see that it uses more than just neural networks (it uses search trees), and that is closer to AI than a plain neural network. $\endgroup$ – Dr. Snoopy May 16 '20 at 16:22

There's some useful information in your description, but that's just a very vague description of how neural networks with sigmoid activation functions are trained.

Moreover, there are many other AI systems apart from neural networks (such as support vector machines, expert systems, etc.), which, of course, I cannot exhaustively list here.

Is my understanding of how AI works correct?

I would say it's not completely incorrect, but, as I said, it's a very vague description and it only refers to a subset of techniques in the AI field. With that description, no newbie would probably understand how neural networks are really trained, apart from knowing that you will train them iteratively and there are sigmoids involved.

  • $\begingroup$ He claimed to be working in AI, so I guess, he might know about it. I'm asking to see if my understanding is correct. $\endgroup$ – noviceFedora May 15 '20 at 17:35
  • $\begingroup$ @noviceFedora I am also studying and working in AI, so I know what I am talking about. You didn't say what that user told you about your description. $\endgroup$ – nbro May 15 '20 at 18:02
  • $\begingroup$ I think he said "pretty much" as in pretty much what I said happens. $\endgroup$ – noviceFedora May 15 '20 at 22:07
  • $\begingroup$ @noviceFedora As I say in my answer, what you say is not incorrect, but it's very very vague! A newbie would require a better explanation and would not really understand how neural networks are trained with your explanation. But you're on the right track. You just need to study more now! Anyway, if your friend said that your explanation is correct, he probably doesn't know much about the topic or maybe he's just careless. Training a neural network is quite more complex than what you said. There are many other details. $\endgroup$ – nbro May 15 '20 at 22:13
  • $\begingroup$ Actually he didn't say pretty much, I misremembered it, he actually didn't answer anything. You can view the discussion here: math.stackexchange.com/questions/3658539/… $\endgroup$ – noviceFedora May 16 '20 at 1:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.