Machine learning in trading: theory, models, practice and algo-trading - page 1816

 
Uladzimir Izerski:

Well, history can change at any time and the neural network will not change the nature of the price movement. Don't even dream). And you can wait years. Sync by honeybunny

I thought that there will be a meaningful dialog, but I was wrong...

 
mytarmailS:

Please don't take the word optimization as a synonym - adjusted period at stochastic.


When you analyze your patterns and try to identify quality entry points for say a buy, sell and garbage pattern...

What do you do? You minimize class error(this is optimization), and there are three classes: "buy" , "sale" This is aclassification problem in machine learning.

When you have broken down market situations into patterns and given them numbers, that is the task of clustering in machine learning.

When you find the optimal number of patterns, that's optimization as well.


You see, you have been doing machine learning for a long time and don't even know it, only with non-optimal tools.

You are contradicting yourself.

And how do you know it's a hint of the future? Probably because you've seen similar before, and before it's kind of like history )

Yes, I agree. You got me all worked up. You can't know the future without knowing history. I repent))

 
Evgeny Dyuka:

I thought there would be some kind of meaningful dialogue, but I was wrong...

The dialogue is just beginning, maybe not just with you. Stay tuned. Many interesting things to come.

 
Uladzimir Izerski:

Well, history can change at any time and the neural network will not change the nature of the price movement. Don't even dream). And you can wait years. Hee-hee

The meaning of neural networks is not in memorizing the history, but in generalizing it, in order to produce the correct result when new data is fed. This is what needs to be understood. If a network has a generalizing ability, it will work on new data which is not known to it. Work correctly I mean. So your knowledge of neronal networks is extremely superficial.
 
Mihail Marchukajtes:
The point of neural networks is not to memorize history, but to generalize it in order to give the right result when new data is presented. This is exactly what needs to be understood. If a network has a generalizing ability, it will work on new data which is not known to it. Work correctly I mean. So your knowledge of neronal networks is extremely superficial.

So I am not an Ace in NS. I came to learn from you. And my opinion and attitude to the NS I am not yet forbidden to express.

 
Uladzimir Izerski:

So I'm not an ace in the NS. I came to learn from you. And my opinion and attitude to NS I am not forbidden to express yet.

Yes, I'm sorry for being so abrupt. Well done for coming!!!! There are a lot of experienced users of the NS
 
Mihail Marchukajtes:
Yes, I'm sorry for being so abrupt. Good for you for coming!!!! There are a lot of experienced users of NS

I will not refuse new ideas of NS, but I am more interested in AI.

 
Uladzimir Izerski:

I won't say no to new NS ideas, but I'm more interested in AI.

First of all, give me a definition of AI

 
mytarmailS:

First, give a definition.

I have not dug into the best meanings, but the essence is the same. My meaning is that AI is a software product with the ability to understand the current situation and see the future.

 
Uladzimir Izerski:

I have not dug into the best meanings, but the essence is the same. My meaning is that AI is a software product with the ability to understand the current situation and see the future.

ok

what is situational awareness ?

What is perspective ?


You're going to write code, not play with words, right?


You can't tell the computer, give it a computer, understand the situation.)

And he'll give you a sack of bones)

Reason: