Machine learning in trading: theory, models, practice and algo-trading - page 1601

New comment
 
Alexander_K2:

:))) This is for all those who are suffering. To have an understanding that there is nothing on "raw" data and no mathematical device will cope with usual OPEN/CLOSE or ticks. That the stages of data preprocessing and thinning are necessary for any TS.

And the way to prepare market quotes for the analysis - everyone is free to be creative. Temporal :)))


Theoretically it is clear what has to be done

But so far it is absolutely unclear how to implement it in practice

 
Renat Fatkhullin:

Preparing full-fledged machine learning with Python support right in Metatrader: https://www.mql5.com/ru/forum/306688/page29#comment_14728528

This is a very important priority for us in the development of algotrading in MT5.

Please take a look at it.

Use Keras and Tensorflow 2 libraries.
 
Mihail Marchukajtes:
Hello all!!! Yesterday I had an incident that turned my life upside down. I was offered to quit my job and I said yes, so the streaming is not far off, but I have already bought a Goproscope, so I do not need to ask anyone with camera. So, expect :-) Soon, everything will be, we are approaching the promised 3 years :-)
But it would not be advisable to hope that Forex will fill your pockets
 

This is the second year I've periodically looked at the topic, maybe I missed something.

Question. Who was able to implement in practice all that is discussed here, are there results or all is just talk?

For example, at least something like that, I'm not asking about account monitoring.

GBPUSD,M1: FinalTraning... Test!!! in_type[3]: Long, in_profit[3]: -7.9, LongProfit: -17.36 %, LongLost: 99.76 %, ShortProfit: -2.07 %, ShortLost: -5.17 %
GBPUSD,M1: FinalTraning... in_type[3]: Long, in_profit[3]: -7.9, LongProfit: -0.83 %, LongLost: 99.70 %, ShortProfit: -3.75 %, ShortLost: -3.05 %
  
GBPUSD,M1: FinalTraning... Test!!! in_type[2]: Short, in_profit[2]: -1.8, LongProfit: -1.41 %, LongLost: 9.42 %, ShortProfit: -24.48 %, ShortLost: 99.69 %
GBPUSD,M1: FinalTraning... in_type[2]: Short, in_profit[2]: -1.8, LongProfit: -1.77 %, LongLost: 3.75 %, ShortProfit: -7.11 %, ShortLost: 99.38 %
  

GBPUSD,M1: FinalTraning... Test!!! in_type[1]: Long, in_profit[1]: 10.2, LongProfit: 98.06 %, LongLost: 37.42 %, ShortProfit: 15.52 %, ShortLost: 2.40 %
GBPUSD,M1: FinalTraning... in_type[1]: Long, in_profit[1]: 10.2, LongProfit: 99.15 %, LongLost: -6.27 %, ShortProfit: 6.46 %, ShortLost: -3.34 %
  
GBPUSD,M1: FinalTraning... Test!!! in_type[0]: Long, in_profit[0]: 10, LongProfit: 99.01 %, LongLost: 1.26 %, ShortProfit: 0.31 %, ShortLost: 5.54 %
GBPUSD,M1: FinalTraning... in_type[0]: Long, in_profit[0]: 10, LongProfit: 99.01 %, LongLost: 1.26 %, ShortProfit: 0.31 %, ShortLost: 5.54 %
 
Mihail Marchukajtes:
Hello everyone!!! Yesterday an incident happened that turned my life upside down. I was offered to quit my job and I said yes, so the streaming is not far off, but I already bought a Goproscope, so I don't need to ask anyone to help me with the camera. So, expect :-) Soon, everything will be, we are approaching the promised 3 years :-)

The phrase "I was asked to resign and I agreed," i.e., were you fired?

 
Alexander_K2:
For leisure reading:

I didn't see anything new there, unfortunately

and the choice of predictor, again unfortunately, may lead to the fact that the melting of ice in africa depends on the age of J.Lo

strangely enough, I have come to the conclusion that predictor selection is not a science, but an art after all

 
Boris:

strangely enough, I have come to the conclusion that choosing a predictor is not a science, but an art after all

Of course it's an art.

Again, you have to have a clear idea, like, "Yes, I know how to make money on such and such a process. And, then to take it out, to extract that process from the marketplace. Go to a subprocess that has its own time tau=f(t), where t is our normal astronomical time.

 
Alexander_K2:

Of course, art.

Again, you have to have a clear idea, like, "Yes, I know how to make money with this process. And, then you have to take it out, separate that process from the market process. That is, go to a subprocess that has its own time tau=f(t), where t is our usual astronomical time.

everything is so or nearly so, on "pure market time" it is really sad

 
Alexander_K2:

Of course, art.

Again, you have to have a clear idea, like, "Yes, I know how to make money with this process. And, then you have to take it out, to separate that process from the market process. That is, go to a subprocess that has an intrinsic time tau=f(t), where t is our normal astronomical time.

..... And then to understand that it is possible to "fish out" till infinity, but one still has to trade on the market process, not on the "fish out" and to merge in a terrible agony

 
Dmitry:

..... And then realize that you can "fish it out" to infinity, but you still have to trade on the market process, not on the "fish it out", and merge in terrible agony

Hahahaha.... Exactly !)

1...159415951596159715981599160016011602160316041605160616071608...3548
New comment
Reason: