Machine learning in trading: theory, models, practice and algo-trading - page 2418

 
mytarmailS:
Have you tried it?
An acquaintance tried it (we just communicated), he says that the results are at the boosting level.
 
mytarmailS:
Yuri do not believe in a fairy tale... all I can say is...

No fairy tale required. I have no need for a fairy tale, just respect for my customers. The Expert Advisor has started to steadily leak, a lot of requests in the comments, personal messages. In this regard, I should mention that Maksim has gone into the shadows, he hasn't written anything and hasn't answered anyone. He has been discussing technical questions here quietly. He takes offense even when a remark is made.

Any algorithm can break, it is a fact.

Respectful treatment in this case - is the appeal to their customers with a message that something is not working as intended, let's all disable the EA, now I'll figure out what went wrong, I'll do everything, I'll post updates, everything will be fine.

Acknowledging an error and trying to solve it is a sign of professionalism, rather than turning a blind eye and ignoring the problem

 
YURY_PROFIT:

It doesn't require a fairy tale. Just a respectful respect for my customers. The Expert Advisor started to leak steadily, a lot of requests in the comments, personal messages. This is why the analytical analyzer hasn't yet started to analyze the real trading robot, but it has not responded. He takes offense even when a remark is made.

Any algorithm can break, it is a fact.

Respectful treatment in this case - is the appeal to their customers with a message that something is not working as intended, let's all disable the EA, now I'll figure out what went wrong, I'll do everything, I'll post updates, everything will be fine.

Recognizing an error and trying to solve it is a sign of professionalism, rather than turning a blind eye and ignoring the problem

So far so good, the trigger hasn't worked. Let's end the discussion here, otherwise we'll both get banned. Write in the comments, I've been on the road for a few days, just got to the site and responded immediately.
 
Maxim Dmitrievsky:
An acquaintance of mine tried it (we just talked), he says that the results are at the boosting level.
I see, well, that's what the article says... I won't even try it)
 
mytarmailS:
I see, well, that's what the article says... I won't even try it)
2019 algorithm, maybe they came up with a newer one
 
Maxim Dmitrievsky:
So far so good, the trigger has not worked. Let's finish the discussion here, otherwise we'll both get banned. I've been on the road for a few days and just got to the site, so I wrote back.

Maxim, if you were anywhere else to respond, I would not write here. Let them ban you, maybe you'll take care of your adviser!

 
YURY_PROFIT:

It doesn't require a fairy tale. Just a respectful respect for their customers. The Expert Advisor started to leak steadily, a lot of requests in the comments, personal messages. This is why the analytical analyzer hasn't yet started to analyze the real trading robot, but it has not responded. He takes offense even when a remark is made.

Any algorithm can break, it is a fact.

Respectful treatment in this case - is the appeal to their customers with a message that something is not working as intended, let's all disable the EA, now I'll figure out what went wrong, I'll do everything, I'll post updates, everything will be fine.

Admitting an error and trying to solve it - this is a sign of professionalism, rather than turning a blind eye and ignoring the problem

the algorithm cannot break, the algorithm can only become defective over time, because it has that kind of fate

so if the force majeure question of stability and stability is not solved, it is better not to implement it

And you were right about the fairy tale above, over time you will understand the depth of those words

 
Maxim Dmitrievsky:

This alternative is proposed, it seems that normalization is not necessary. I have not tried it yet.

Another thing to note is that tabular data and timeseries are not exactly the same thing. That's why there may be a glitch on the timeseries.

https://towardsdatascience.com/tabnet-e1b979907694

From the article, I liked one thought in the conclusion:

As AI regulation tightens, understanding how our models work will become even more important in the future.

I think it's true. Or else there was already a news report that a woman was fined for going outside under quarantine.
The NS identified that the person in the photo was most likely her.
And you can't argue with that... The NS identified her and that's it, it's impossible to explain how the NS decided this.
I do not know how it ended ...

 
YURY_PROFIT:

It doesn't steer at all, only losses.

- I want a car to win the race for sure!

- here's a Ferrari for you.

---

- your car is rubbish and it sucks. I didn't make it to the finish line, the car is wrecked, there are bodies everywhere... I could hardly get away from the police to say "the car sucks!!!"

- idiot! Tour de France ...

PS / only read the thread, but here I could not help but say

 
elibrarius:

I liked one thought in the conclusion of the article:

I think it's right. Otherwise, there was already a news report that a woman was fined for going outside under quarantine.
The NS identified that the person in the photo was most likely her.
And you can't argue with that... The NS identified her and that's it, it's impossible to explain how the NS decided this.
I do not know how it ended...

It's the same with trading. I have done it, checked it with new data, why it stopped working later is unclear. It often makes no sense to take apart the whole model to understand it
Reason: