What types of problems can neural networks solve? (except optical character recognition)

Question

Neural networks seem really cool, but what types of problems can they solve?

Answer 1

They are well versed in the pattern, not just characters. So they can tell you that this is a picture showing a man or woman, or this picture shows a person xy. Please note that they do not give an exact answer, but only the probability. Also signal patterns such as audio signals. For example, sound x sounds like sound y, etc.

They cannot tell you what the answer is - this is the xy formula, because small changes in the formula can give a completely different answer.

Answer 2

Neural networks can deal with a lot of different problems. For example, in our case, we used them to successfully reproduce stresses, forces, and eigenvalues ​​in the loaded parts (for example, in finite element analysis problems). In technology and computer science there is a whole world where they can be successfully applied.

Answer 3

I would say that there is any problem when ai should learn from previous experience. for example, pandoras box or last.fm, where the application finds out what kind of music the listener is listening to, offering suggestions for voting when listening up or down.