Using Machine Learning to trade Forex succesfully

Hello there,

It’s about 1 year I’m trying to build a succesfully trade system based on machine learning tools such as Neural Network, regression, Random Forest and so on…
To do that I’ve been using tools such as Matlab, R-Studio, Spark on Bigdata Technologies but I really no had good results.

I’ve been using as ‘predictors’ #Pips of body, low and high candles on different timeframe.
Have you got any better experience in trying to training models?

I was thinking to use macroeconomic data to predict long term trade which maybe is easier than short term.

Thanks for any advice.

1 Like

Building machine learning strategies that can obtain decent results under live market conditions has always been an important challenge in algorithmic trading.

1 Like

Did you make any progress on this?

Hello,

Yes, probably I did… but can you believe that after 2 years I am still experimenting?

Wow, it’s great to see that you continued experimenting for 2 years. I, too, have been practicing on similar lines and a course on Machine Learning by Quantra really helped me. It cleared my concepts on ML, while at the same time explaining me how to apply it to trading. I think you should check out this course.

hiya,

how do you connect MATLAB, R-Studio, and Apache-~Spark to MT4 or MT5? What’s the need for R-Studio? Why MQL4 or MQL5 can’t do it?

Thnx

Hi Campione,

so there are at least two ways:
One is to use TCP Sockets to enable communications
One is to use the file system as comunication channel

I used both options for connecting MT4 to Python.

1 Like

Congratulation on your effort. Making such a model is huge challenge especially in contemporary market. There are so many variables, uncertainties and especially connections that should be included in one such model, that makes it questionable if even neural networks can support it