I am the title
Small premise: first you have to be familiar with the original form of the perceptron
Knowing that the author with the second highest vote understands. However, Zhihu has some controversy over the removal of the sum symbol in the parameter update with the highest vote. The author said it was computational thinking, and the commentary said it was stochastic gradient descent.
Attach the link, I am more optimistic about the answer with the second highest vote.
The second highest ticket (you can look at this first)