Discussion of article "Using MetaTrader 5 Indicators with ENCOG Machine Learning Framework for Timeseries Prediction" - page 6
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Awesome article investeo!
Just one problem, every time I run the EA on the simulator I get the following error...
2012.08.07 20:10:36 Core 1 2012.01.01 00:00:00 Access violation read to 0x00000000
I have followed the suggestions on this page. I still get the error.
Please help?
Excellent article!! It's very clear for a beginner like me :)
I'm having the same problem with the attached encog version on Windows 10 X64.
Could anyone solve this? Any suggestions?
Thanks
This article is well written and definitely inspiring. I saw this article translated in Russian and Chinese, impressive!
Anyway, is there anyone, before congratulating with the authors, who tried the entire codes? I have serious doubts that we are really close to random guessing, as expected with Forex.
I run a LSTM as mentioned in the article, focussing not on regression (guessing the next values), but on binary classification: if the max_close in the next 5 future steps is greater than the actual, than class_0 otherwise class_1. In few words, I ran an LSTM to predict if the close in the next future goes up or down. My precision is close to random guessing and I am pretty sure also the precision of the RNN of this article goes close to random guessing. I am really curious about the performances of the RNN of this article. Can the authors share the confusion matrix.
I also ran a fuzzy classifier just classifying the future close the same as the current: it is not definitely a prediction! With that I was able to reach similar profit of this article.
In conclusion, I suspect that the results of the RNN of this article are really close to random guessing and I would like to ask the authors to share with us the direct performance of their RNN.
Anyway, the article is inspiring for further research.
Cheers,
Marco
Looking at the backtest REPORT. Not a very good EA. neither the number of consecutive losses, nor the profit ratio. Although it is profitable, but not very stable.
The conclusion is that neural network algorithms for predicting time series are not a panacea, and with the current state of the art, they are far from the level of artificial intelligence.
However, the idea is good, especially with the three indicator values as the input source for the neural network.
I believe that the creator could explain better where to save each of the files, the EA did not work in my backtest.
I followed all the steps you provided.
Can you be more detailed about where exactly to save each file and where each file is located?
Any idea on how to fix this issue?
2021.03.23 12:03:27.962 2020.10.01 00:00:00 Access violation at 0x00007FF9FE2688C2 read to 0x0000000000000000
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D6F0 4881ECD8000000 sub rsp, 0xd8
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D6F7 488B0572E72500 mov rax, [rip+0x25e772]
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D6FE 4833C4 xor rax, rsp
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D701 48898424C0000000 mov [rsp+0xc0], rax
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D709 488364242800 and qword [rsp+0x28], 0x0
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D70F 488D05DAFFFFFF lea rax, [rip-0x26]
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D716 83E201 and edx, 0x1
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D719 894C2420 mov [rsp+0x20], ecx
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D71D 89542424 mov [rsp+0x24], edx
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D721 4889442430 mov [rsp+0x30], rax
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D726 4D85C9 test r9, r9
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D729 744C jz 0x7ffa7439d777
2021.03.23 12:03:27.966 2020.10.01 00:00:00
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D72B B80F000000 mov eax, 0xf
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D730 488D4C2440 lea rcx, [rsp+0x40]
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D735 443BC0 cmp r8d, eax
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D738 498BD1 mov rdx, r9
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D73B 440F47C0 cmova r8d, eax
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D73F 4489442438 mov [rsp+0x38], r8d
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D744 49C1E003 shl r8, 0x3
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D748 E82A470600 call 0x7ffa74401e77 ; SetProcessDynamicEnforcedCetCompatibleRanges (kernelbase.dll)
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D74D 488D4C2420 lea rcx, [rsp+0x20]
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D752 48FF15AF231900 call qword near [rip+0x1923af] ; UnhandledExceptionFilter (kernelbase.dll)
2021.03.23 12:03:27.966 2020.10.01 00:00:00 crash --> 00007FFA7439D759 0F1F440000 nop [rax+rax+0x0]
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D75E 488B8C24C0000000 mov rcx, [rsp+0xc0]
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D766 4833CC xor rcx, rsp
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D769 E8D2090600 call 0x7ffa743fe140 ; RemoveDllDirectory (kernelbase.dll)
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D76E 4881C4D8000000 add rsp, 0xd8
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D775 C3 ret
2021.03.23 12:03:27.966 2020.10.01 00:00:00
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00007FFA7439D776 CC int3
2021.03.23 12:03:27.966 2020.10.01 00:00:00
2021.03.23 12:03:27.966 2020.10.01 00:00:00 00: 0x00007FFA7439D759
2021.03.23 12:03:27.966 2020.10.01 00:00:00
Use Mine @Valentin petkov. i am using encog 3.3 . i hope can help you ..
Hi ryuga68. Class TemporalWindowCSV does not exist in encog >= 3.3 ??