Adaptive EA for MT5
- Experts
- Svetlana Visnepolschi
- Versione: 1.6
- Aggiornato: 17 gennaio 2022
- Attivazioni: 16
It is a trainable trading advisor that you can adapt to trade on any symbol. The robot is suitable for trading any currency pairs, metals, oil or cryptocurrencies. To learn, just run the robot in the tester. At the first pass of the tester, the neural network is trained performing random transactions. During the second pass, you can see the trading results using your trained neural network. Before using the advisor, you can watch the tutorial video.
Parameters
- Lots size - size for orders;
- Lots risk - size for orders as a percentage of the deposit;
- Grid risk - grid order size multiplier;
- Profit for grid orders - total take profit of the grid;
- Step for grid orders - step between grid orders;
- Orders count maximum - grid orders count;
- Grid reverse trading - reverse grid;
- Margin level minimum - minimum margin level, upon reaching which the order will be closed and it becomes impossible to open new orders;
- Take profit - take profit size in pips;
- Stop loss - stop loss size in pips;
- Trailing stops - size of the trailing stop in pips;
- Max spread - maximum spread;
- Buy orders are restricted - prohibits opening buy orders;
- Sell orders are restricted - prohibits opening sell orders;
- Process adaptation on history bars - the number of bars participating in the training sample;
- Process adaptation every time - overwrites the neural network in any case, even with a forward test;
- Ignore Pattern Errors - ignore errors (for example, gaps) on the chart while working;
- Use VPS Migration - this parameter must be enabled before migrating to VPS from MQL (experimental mode, only one running copy of EA is available);
- Monday Trade Session - Sunday trade Session - trade start and end time (do not trade if the value is empty);
- Magic number - orders magic number;
- Comments - comments for orders opened by the advisor;
- Slippage - maximum slippage level;
- Show info panel - shows informational dashboard;
- Adaptation Folder - directory for trained neural networks;
The Back test and Forward test (MT5) are perfect since the Forward test (with 1/2, 1/3, 1/4 and custom) still gives very good results. With some work on adjusting the take profit and stop losses (also max spreads+Trailing Stops), the test gives a very good result on saveral other pairs as well. I have used this EA on 3-4 pairs on low latency VPS (1-4ms) in the past 5 days and the results are great for me. Thank you to the author of this EA and hope it can be further improved to be one of the best EA out there.