What to feed to the input of the neural network? Your ideas... - page 4

 
Dmitry Fedoseev #:

So you have not yet understood what a neural network is, why it is needed and how it works.... How can you talk such rubbish? How can you get by with a stochastic formula? Well, it's utter nonsense, absolute ignorance, but so much poncey.

The beginning is correct (highlighted) ...

 
Sergey Chalyshev #:

The beginning is right (emphasis added) ...

I have a very vague guess that you wanted to communicate something .... but you didn't.

 

And the rest of it is rubbish.

As they say :

you are not fit to be friends, but you are not fit to be musicians.)

Is it so hard to understand what a neural network is?

 
Dmitry Fedoseev #:

I have a very vague hunch that you wanted to communicate something.... but you didn't.

Dimitri, why do you need a neural net?

Can you at least answer that for yourself?

Do you know what a PLM is?
 

Eh... typical forum with trolls and goofballs.... I gave a clear answer with a leading question. Thought they'd come round... Well, well, well...

Maxim Kuznetsov #:

and that's what the NNs are for?

Dmitry Fedoseev #:

Masterpiece! You can immediately feel the depth of immersion in the topic and the power of knowledge.

 
Dmitry Fedoseev #:

I can. An outboard boat motor - of course I know - an outboard. It's suspended from a pole cut out of wood. What are you interested in?

Joker)))

Outboard boat motor - cool).

I am not interested.

In your person addressed to all fans of neurosystems, nothing personal.

PLM - the same neuroset (not fishing net), is only able to memorise what you teach it.

And if you can't trade yourself, you can't teach it anything. Neural network is not able (physically) to generate new ideas.

And in the market nothing repeats for a long time, against a scrap there is always another scrap).

 
When AlgLib is overpowered, then throw on my patches for Pytorch - everything on GitHub is open and without SMS.
 
Sergey Chalyshev #:
Do you know what a PLM is?

That's a very good analogy, by the way!

 
Что подать на вход нейросети?

What is great about neural networks - you input anything you want, and the output is what you have learnt, and no more.

Feed in and out deals from the top signals.

But the result will always be the same, and everyone knows what it is.

 
Sergey Chalyshev #:

Joker)))

Outboard boat motor - cool)

I'm not interested.

In your person I addressed to all fans of neuroses, nothing personal.

PLM is the same neural network (not a fishing net), it can only memorise what you teach it.

And if you can't trade yourself, you can't teach it anything. Neural network is not able (physically) to generate new ideas.

And in the market nothing repeats for a long time, against a scrap there is always another scrap).

This outboard is nothing like a neural net. Nothing in common at all.

If we proceed from the postulate that nothing repeats itself in the market for a long time, then you should not apply any method at all.