I've never seen this notation before, so I'm not sure where to start learning about it:


I got it from here: http://neuralnetworksanddeeplearning.com/chap1.html#the_architecture_of_neural_networks (about half way down)

It looks like he's adding two partial derivatives, one for C with respect to v1 and one for C with respect to v2. That makes sense sort of, but why is he multiplying the derivative against the change in v1 and the change in v2.


For a differentiable real valued function of two variables $F(x,y)$, a linear approximation of the change in $F$ when $(x,y)$ moves to $(x+\Delta x,y+\Delta y)$ is given by:

$$\Delta F\approx \frac{\partial F}{\partial x}\Delta x+\frac{\partial F}{\partial y}\Delta y$$

where the partial derivatives are calculated at $(x,y)$. Is it more familiar now?

  • $\begingroup$ Ahhh, linear approximation. Got it. Just saying that the rate of change * change = actual change. Thanks!!! $\endgroup$ – Blaze Nov 15 '15 at 17:45

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.