Machine learning in trading: theory, models, practice and algo-trading - page 853

 
Mihail Marchukajtes:
Hey listen, smart guys!!!! Don't you have the guts to solve my problem?

Why do we need your tasks? We have enough of our own))

 
Mihail Marchukajtes:

You know what you're saying?????

The pot doesn't cook at all?

 
Yuriy Asaulenko:

Why do we need your tasks? We have enough of our own.)

you cardboard idiot.... This task is the KEY!!!!

 
Mihail Marchukajtes:

You cardboard fool.... This task is KEY!!!!

That's a matter of opinion. It might be key for you. So decide, dear.

 
Maxim Dmitrievsky:
I just cry when I see that you keep picking up the importance of predictors for some pieces of market history. Why? It's a profanation of statistical methods.

Why? your question sounds exactly like that doesn't it. ???? I'll tell you why.....

Because it is enough to choose from the variety of input data exactly what we were able to determine with the help of statistical methods and calculation of entropy of VI, etc. We've given the input exactly what is important for the output. That is, this preprocessing only concludes that there is some relationship between input and output. Just that!!!! Yes in these five inputs there is a dependence to the output. And what exactly is this dependence? This is what the AI system and optimizers are looking for. And after the obtained models we make an estimate as shown in the figure and find the exact model that works. So which model in my figure is working A? Б ? C? Or D?

Neural network is not a dumpster, what you put in it, and you reap. I'm surprised you didn't know that until now.....

 
Yuriy Asaulenko:

That's a matter of opinion. You might be the key one. That's up to you, honey.

It's key for you, too, you don't really get it. ???? That's your problem. You don't want to hear what you are told....

 
Mihail Marchukajtes:

Why? Your question sounds exactly like that doesn't it. ???? I'll tell you why.....

Because it is enough to choose from the variety of input data exactly what we were able to determine by using stat methods and calculating the entropy of VI, etc. We've given the input exactly what is important for the output. That is, this preprocessing only concludes that there is some relationship between input and output. Just that!!!! Yes in these five inputs there is a dependence to the output. And what exactly is this dependence? This is what the AI system and optimizers are looking for. And after the obtained models we make an estimate as shown in the figure and find the exact model that works. So which model in my figure is working A? Б ? C? Or D?

Neural network is not a dumpster, what you put in it, and you reap. I'm surprised you didn't know that until now.

you're only looking at importance for the current piece of the graph, the patterns of which are changing on the OO, just like importance

you'll sift out only the trash through importance

I wonder how hard it is to understand after all these years
 
It is quite possible that you already have systems capable of obtaining adequate models, and sometimes you get them. You consider it an accident only because you do not know how to search or choose from all variety of models built by your own AI. After all the main task in AI is not getting a model, but the choice at their variety or repeated training of the same training file...... Live and learn...
 
Mihail Marchukajtes:

So it's key for you too, don't you really get it???? That's your problem. You don't want to hear what you're being told....

For me? I've already solved the problem. Now I'm thinking of something else to do. Python or R. I have no new ideas yet.

 
Maxim Dmitrievsky:

you only look at importance for the current slice of the graph, the patterns of which change on the OOS, just like importance

Through importance you'll sift out only the outright junk

I wonder how hard it is to understand after all these years.

That's right, you're thinking correctly, So the task of the AI is precisely in the non-stationary series which has a floating pattern. The task of the AI is to maintain performance when this dependence runs away, at least for an insignificant time, but enough to make money. After all, the regularity does not change by leaps and bounds. In place of the main, first entry, there is another one, but the main one is still in the set, and here it is the AI that takes upon itself the load of holding the line, as they say. That's why in the first month of a futures contract, you have to train very often, especially when the market does not know where to go. Looking at the Vtrite, I can see this pattern dancing. But in the middle and at the end of the futures the market usually becomes more orderly and one entry dominates for a long time.