Machine learning in trading: theory, models, practice and algo-trading - page 1030

New comment
 
Renat Akhtyamov:

Let's say, for the Eve, I spoke yesterday about the level of 1.1685. This is the future price level.

If you have the same level in your stash, respectively my and your strategies are the same.

And recently, in one of the threads, on July 12, I said this:

"Example - the eva will be at 1.1712, 1.1744, 100%."

If you have the same level in your puzzle, then my strategy is the same...

....
 
Renat Akhtyamov:
....

Yes yes I remember your predictions, the first time I tilted, but the second time I took my

I don't know if it's better to have a bird in the hand...

 
So will there be any ideas on how to teach neuronics levels? Or did I raise this question in vain?
 
mytarmailS:
So there will be some ideas how to teach the neuron levels? Or I raised this question in vain

For a long time I've been nurturing the idea to train a model by market profile. But so far I do not have enough experience to implement it.)

And additionally net price increments to predictors.

 
mytarmailS:
So will there be any ideas on how to teach neuronics levels? Or did I raise this question in vain?
And how, in your opinion, should learning levels differ from learning on High and Low in bars?
 
Ivan Negreshniy:
And how do you think learning levels should differ from learning on High and Low in bars?

If in general, not much, just looking for something else

 
mytarmailS:

If in general, not much, just looking for another

if you look systematically, you can try to formalize all possible object types for predictions at once, and at the same time possible predictor types as well.
 

I don't know if it's true or not, but it's interesting to listen to

https://www.youtube.com/watch?v=s0MXYOyd_LA

 
Ivan Negreshniy:
if you look systematically, you can immediately try to formalize all possible types of objects for predictions, at the same time possible types of predictors.

The only type of object for prediction is the level (those bounce of the price in the future)

Questions:

1) how to teach the network to rebound if the rebound can be in 5 minutes or an hour or the day after tomorrow, I remind you this is not the next increment to predict

2) how to measure the error with such a target (rebound) then the new data

I don't have an answer yet(!

 
mytarmailS:

The only type of object for prediction is the level (those bounce of the price in the future)

Questions:

1) how to teach the network to rebound if the rebound can be in 5 minutes or an hour or the day after tomorrow, I remind you this is not the next increment to predict

2) how to measure the error with such a target (rebound) then the new data

i don't have an answer yet((

"Bounce" is an abstraction and you can think of many more such images, such as "breakout", but the result will still be a signal to buy or sell, in 5 minutes an hour or the day after tomorrow.
1...102310241025102610271028102910301031103210331034103510361037...3498
New comment
Reason: