Machine learning in trading: theory, models, practice and algo-trading - page 2185

 
mytarmailS:

Does anyone know the technology to represent several values by one with the possibility of converting back to several

with little or no loss of precision

There's no easy one. All combinations in the dictionary and hash keys

Поиск - MQL5.community
Поиск - MQL5.community
  • www.mql5.com
Поиск выполняется с учетом морфологии и без учета регистра. Все буквы, независимо от того, как они введены, будут рассматриваться как строчные. По умолчанию наш поиск показывает страницы...
 
Valeriy Yastremskiy:

there is no simple one. All combinations in the dictionary and hash keys

maybe autoencoder...

 
mytarmailS:

maybe an autoencoder...

Yes, hash with NS is easier to get simpler and more stable).

 
By the way, yeah, how's the pee-pee? Who had the longest? The fact that Rinat is beating the martin at once it is clear from the curve of funds. But what will respond to this other competitor Maksimka?
 
Mihail Marchukajtes:
By the way, yes, how was the pussy? Who had the longest? The fact that Rinat kicks a martin immediately understandable by the curve of funds. But what will respond to it the other contestant Maksimka?

Makhimka will not lose. He took a deposit with a large margin, so as not to lose.

 
Evgeni Gavrilovi:

Who can tell me why the MT5 is not installed in Colab?

command: !pip install MetaTrader5

Gives two errors - Could not find a version that satisfies the requirement MetaTrader5==5.0.33 (from versions: none)

No matching distribution found for MetaTrader5==5.0.33

This question should be asked in another thread

MetaTrader 5 Python User Group - как использовать Python в Метатрейдере
MetaTrader 5 Python User Group - как использовать Python в Метатрейдере
  • 2020.11.20
  • www.mql5.com
Мы готовим модуль MetaTrader 5 для Python...
 
Vladimir Perervenko:

Do you know how to represent several values by one with the possibility of converting back to several,

with minimumloss of accuracy or without


For example with the same umap you can compress data into one vector, but there is no inverse function


 
mytarmailS:

You don't know how to represent several values by one with the possibility of converting back to several,

with minimal or noloss of accuracy.


For example you can use umap to compress data into a single vector, but there is no inverse function


https://www.youtube.com/watch?v=qN3n0TM4Jno&ab_channel=VenelinValkov

https://www.youtube.com/watch?v=H4J74KstHTE&ab_channel=VenelinValkov
Time Series Anomaly Detection Tutorial with PyTorch in Python | LSTM Autoencoder for ECG Data
Time Series Anomaly Detection Tutorial with PyTorch in Python | LSTM Autoencoder for ECG Data
  • 2020.03.19
  • www.youtube.com
Use real-world Electrocardiogram (ECG) data to detect anomalies in a patient heartbeat. We'll build an LSTM Autoencoder, train it on a set of normal heartbea...
 
Maxim Dmitrievsky:

https://www.youtube.com/watch?v=qN3n0TM4Jno&ab_channel=VenelinValkov

https://www.youtube.com/watch?v=H4J74KstHTE&ab_channel=VenelinValkov

You can do it through the pca

https://stats.stackexchange.com/questions/229092/how-to-reverse-pca-and-reconstruct-original-variables-from-several-principal-com


But if you restore only one component at a time, the losses are too big (

And in the trailer it does not matter what to restore with and that is bad (

But we will try!

 
Uladzimir Izerski:

Mahimka is not going to lose. He took the deposit with a large margin, so as not to lose.

As I understand it, I'm not talking about the plum, but about the positive result of the balance curve at the end of the week. That's how I understood it.
Reason: