Machine learning in trading: theory, models, practice and algo-trading - page 757

[Evgeny Raspaev]

Mihail Marchukajtes:

Try looking at this

http://neuroshell.noxapredict.com/en_products.htm

Thank you very much))

[Maxim Dmitrievsky]

Has no one started studying deep q-learning yet? no one to discuss it with? :)

By the way, it's a useful community, join in

http://ods.ai/

[Evgeny Raspaev]

Maxim Dmitrievsky:

Has no one started studying deep q-learning yet? no one to discuss it with? :)

By the way, it's a useful community, join in

http://ods.ai/

I do not want to register, what is the essence of the service?

[Maxim Dmitrievsky]

Evgeny Raspaev:

I don't want to register, what's the point of the service?

Well, the community of neural networks, and everything associated with it, mostly python.

[Evgeny Raspaev]

Maxim Dmitrievsky:

Well, the neural network community, and everything related to it, mostly python.

How did you learn about opendatascience? (Name and nickname of the person in slack who told you and called you)*
From whom to write? Maybe from you, maybe some bonuses will come to you

[Maxim Dmitrievsky]

Evgeny Raspaev:

How did you learn about opendatascience? (Name and nickname of the person in slack who told you and called you)*
Who should I write from? Maybe from you, maybe some bonuses will come to you

no there are no bonuses :) you can write anything

[Anatolii Zainchkovskii]

What if we represent a sequence of "photos" of the chart with all the desired indicators as input data? It is necessary to stipulate, of course, that not the photos themselves, but the numbered "pictures".

[Maxim Dmitrievsky]

Anatoly Zainchkovskii:

We got a batch of new ideas and quickly decided to code and test them all)

Anatoly, you also use scaffolding, can you tell me how to save the trained model and load it again?

I can't do it via serialization/Deserialization.

[Anatolii Zainchkovskii]

Maxim Dmitrievsky:

Anatoly, you also use scaffolding, can you tell me how to save the trained model and reload it?

I can't use serialization/deserialization.

I am using it, but so far at a weaker level than you. i have also thought about saving it. i think there is a way to save the trained model to a file as an array. if i'm not mistaken, the output of a trained forest is an array.

[Maxim Dmitrievsky]

Anatolii Zainchkovskii:

I am using it, but so far at a level weaker than you. I was also thinking about saving. in my opinion, there is a variant of saving an array to the file. if I am not mistaken, the trained forest is an array at the output.

there are such fields at the object (Trf - trained forest), and only the last field is an array

maybe they contain all the information, i should check it

 Print(Trf.m_bufsize);
 Print(Trf.m_nclasses);
 Print(Trf.m_ntrees);
 Print(Trf.m_nvars);
 ArrayPrint(Trf.m_trees);

///////////////////////////////////
2018.03.19 19:36:29.690 RF sample (EURUSD,D1)   4730
2018.03.19 19:36:29.690 RF sample (EURUSD,D1)   1
2018.03.19 19:36:29.690 RF sample (EURUSD,D1)   10
2018.03.19 19:36:29.690 RF sample (EURUSD,D1)   2
2018.03.19 19:36:29.699 RF sample (EURUSD,D1)   [   0] 473.00000   0.00000   0.55000 236.00000   0.00000   0.25000  99.00000   0.00000   0.15000  57.00000   1.00000   0.55000  35.00000   1.00000   0.35000
2018.03.19 19:36:29.699 RF sample (EURUSD,D1)   [  15]  28.00000   1.00000   0.15000  21.00000  -1.00000   0.01000   1.00000   0.25000  26.00000  -1.00000   0.02000  -1.00000   0.03000   1.00000   0.45000
2018.03.19 19:36:29.699 RF sample (EURUSD,D1)   [  30]  33.00000  -1.00000   0.04000  -1.00000   0.05000   1.00000   0.75000  45.00000   1.00000   0.65000  43.00000  -1.00000   0.06000  -1.00000   0.07000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [  45]   1.00000   0.95000  55.00000   1.00000   0.85000  53.00000  -1.00000   0.08000  -1.00000   0.09000  -1.00000   0.10000   1.00000   0.55000  77.00000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [  60]   1.00000   0.30000  70.00000   1.00000   0.15000  68.00000  -1.00000   0.02000  -1.00000   0.04000   1.00000   0.45000  75.00000  -1.00000   0.08000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [  75]  -1.00000   0.10000   1.00000   0.75000  87.00000   1.00000   0.65000  85.00000  -1.00000   0.12000  -1.00000   0.14000   1.00000   0.95000  97.00000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [  90]   1.00000   0.85000  95.00000  -1.00000   0.16000  -1.00000   0.18000  -1.00000   0.20000   0.00000   0.35000 144.00000   1.00000   0.45000 122.00000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [ 105]   1.00000   0.25000 115.00000   1.00000   0.15000 113.00000  -1.00000   0.03000  -1.00000   0.06000   1.00000   0.35000 120.00000  -1.00000   0.09000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [ 120]  -1.00000   0.12000   1.00000   0.75000 132.00000   1.00000   0.60000 130.00000  -1.00000   0.15000  -1.00000   0.21000   1.00000   0.85000 137.00000
2018.03.19 19:36:29.700 RF sample (EURUSD,D1)   [ 135]  -1.00000   0.24000   1.00000   0.95000 142.00000  -1.00000   0.27000  -1.00000   0.30000   0.00000   0.45000 194.00000   1.00000   0.55000 172.00000