How is classification done on Support Vector Machines?

Question

Machine learning

Based on the material studied, I realized that the classes are separated by a hyperplane, and those elements that are closer to it are the support vectors. But how does the program determine which class it needs? For example, there are several types of animals, how then to choose the right class? Take all pivot points and compare them to the input?

Answer 1

the initial weights can be randomly distributed, that is, randomly determine which class is in front of you, for the current weights. then anyway you change the configuration of the hyperplane space, adjusting it to the desired result, so that there are more correct answers, through the feedback mechanism)
well, in general, the classical scheme, where the starting values ​​​​do not matter much, the result will still be the same.