Answer the question
In order to leave comments, you need to log in
What is the dependence of the number of A-elements on the number of layers in the Rumelhart perceptron?
As I understood from the Wikipedia article , for the best training of a Rosenblatt single-layer perceptron, we need as many A-elements as there are input data options (training examples)
If the number of stimuli in the space W is equal to n > N (that is, it is greater than the number of A-elements of the elementary perceptron), then there is some classification C(W) for which there is no solution.
It follows that:
in Rosenblatt, the number of A-elements is equal to the number of stimuli
There is an assumption that by using more layers, you can reduce the number of elements in them, that is, the total number of elements in the layers will be less than if you use one hidden layer. This assumption has been successfully used in deep learning technologies and has a rationale.
Answer the question
In order to leave comments, you need to log in
In this case, the easiest way is to hang the event handler using JavaScript methods.
I would do like this:
My implementation can be called far from ideal, but if they throw stones, at least it is argued.
Important ! Familiarize yourself with an attribute like data-* before using it. I found it useful enough for my needs. I hope I helped you.
It is logical to use several containers for a slider and change the classes in both containers when triggering slideChange (depending on your method of creating a slider: ribbon or classes). If you want to use a ready-made solution, you can connect two slick sliders and specify the asNavFor option in one of them ( kenwheeler.github.io/slick slider syncing section).
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question