I didn't know which part of the forum this would be best suited so apologies in advanced!
We recently concluded and released a two year long project attempting to introduce machine learning/an adaptive way of trading through self-adaption for Forex (cfd/spread betting & binary).
I won't go into too much detail here as it would look like spam but its not. Please check out our flow diagram here of our processes: http://efxone.com/EFXONE-System-Flow-Chart.html
The idea is simple; before the algorithms get to work, there is an additional framework layer that alters parameters within the detection algorithms based on market conditions. Conditions are evaluated using scenario based models via the behaviour of the asset. More information on the framework here: http://efxone.com/Framework
The actual algo's use various techniques including Wave analysis, Ichiomku cloud strategies, price action, momentum, order flow and many others. Detailed information can be found on our one page website: http://www.efxone.com
I would really love to hear what you guys think of this. Please ping me across any questions on here or our twitter: www.twitter.com/efx_one - we are also posting daily results of real trades as they happen.
I look forward to connect with some of you for constructive conversations!