Expert Advisors: Support Vector Machine Learning Trader

New comment
 

Support Vector Machine Learning Trader:

This Expert Advisor uses support vector machine learning to analyze historical price data and signal future trades.

Author: Josh Readhead

 
Automated-Trading:

Support Vector Machine Learning Trader:

Author: Josh Readhead

When I download an EA, I don't understand why I can't add it to the graph, I don't even see it in: "Insert->Experts". I only see old EAs !!
 
anyone tested this EA? It is only for testing?
best regards.
Antonio
 

Hi. The code specifies the file "svMachineTool.ex5", where is it?

//+-------Functions to Support Vector Machine Learning tools-------+
//|| The following #import commands import all functions
//|| systems based on the support vector method to be used in this Expert Advisor. 
//|| Note that if you don't import functions here, the compiler will not allow
//| use these functions|
//+------------------------------------------------------------------+
#import "svMachineTool.ex5"
 
pafn61:

Hello! The code specifies the file "svMachineTool.ex5", where is it?

Apparently here:

... (a copy of the tool for training based on the support vector method is available for download in the MQL5 Market) ....
 

That phrase

//|| Please note that the Expert Advisor is intended for demonstration purposes only 
//| tool using the support vector method. It is recommended to make changes in the Expert Advisor,
//| to adapt it to your trading style.
should probably be in the description.
 

Hi Antonio

This expert advisor was written to be used in conjunction with the Support Vector Machine Learning Tool available on the metaquotes market. A free demo version of this can be downloaded from the previous link or from the terminal. The tool is a library of functions that uses support vector machine learning to assess selected indicators as inputs and based on a training data set (generated from historical data) provide signals for entering buy/sell trades. A better description of how the library works can be found on the market page.

This expert advisor ("Support Vector Machine Learning Trader") has been written as an example of how to implement this tool in trading. The code has been documented heavily so feel free to modify as it suits you. 

I have written an article also on how Support Vector Machines can potentially be used in trading decisions, this goes into a bit more detail on how Support Vector Machines work and how they can be implemented in trading. It is called  "Machine Learning: How Support Vector Machines can be used in Trading" and is currently being reviewed by moderators, so it should be published to the site shortly.

I hope this clarifies a bit.

Regards

Cubix 

 

If you are having trouble using this expert advisor with the demo version of the Support Vector Machine Library, try modifying the #import statement at the top of the code (line 45).

It says;

#import "svMachineTool.ex5"

but is should say;

#import "svMachineTool_demo.ex5"

When downloading the demo version, mql5 adds the "_demo" suffix to the file name to distinguish between the demo and full version. If you have a full version of the library, there should not be any issue.

Also note that you can only use the demo version of the library in strategy tester.

- Jasper

 

I have setup another forum to specifically discuss questions/answers around this EA and the Support Vector Machine Learning Tool. The link is https://www.mql5.com/en/forum/9596 

Feel free to visit and contribute your questions/answers... 

 
Help, I bought your library, and first I tried to run an example of your own Expert Advisor with the purchased library and I get the following error every time: 2013.01.13 15:47:54 Core 1 2012.12.28 02:46:00 Training has not been successfully completed : Classify Aborted.

 
rushid:
Help, I bought your library, and first I tried to run an example of your own expert with the purchased library and I get the following error every time: 2013.01.13 15:47:54 Core 1 2012.12.28 02:46:00 Training has not been successfully completed : Classify Aborted.

Actually it is better to write to the author in private

and preferably in his native language

12
New comment