Project Description

Self-optimizing, multi-currency EA based on machine learning.


  • TradePairs – List of currency pairs, separated by commas;
  • Lots – Fixed lot size if 0 percentage of deposit is used;
  • Percent – Lot size as a percentage of deposit;
  • TakeProfit – Take profit in five-digit points;
  • StopLoss – Stop loss in five-digit points;
  • BUY Tolerance – Tolerance to generate a BUY signal;
  • SELL Tolerance – Tolerance to generating a SELL signal;
  • Training – Optimization modes;
  • Training Period – Period size in bars for optimization;
  • Training Pattern – Pattern size in bars for optimization;
  • Training Step – Step in bars through which optimization can be repeated;
  • Magic number – Magic number for identifying advisor orders;

The easiest option for one-time optimization – run the EA with default parameters in the tester on any currency pair and timeframe.

After the end of the test, the EA is automatically optimized for the period specified in the Training Period parameter, saves the trained image to a file and will be ready to go.