The derivative: how much the output moves if you nudge the input a little
The derivative answers one question: “If I tweak this weight slightly, does the error go up or down, and by how much?” That is exactly what training needs. Given the derivative, you can step toward lower error — that is gradient descent. Without derivatives, neural networks could not learn at all.
Content is available with subscription.
Get full access to all courses on the platform for one year with a single payment.
▼
Unlike other platforms that charge per course, here you get everything for one price, and after one year of use there will be no automatic charge for the following year.