You know bout AI, ML, Neural, deep learning it sounds interesting, but no one knows how they really effective or not, we could find some in commercial section some EA for several thousand $
Recently we have very aplaudable interesting situation. On MQL5 for first time a coder released soure file of self-adaptive deep learning strategy.
There is a variable that store all market patterns, in parallel a virtual trading system that open a buy and sell on each candle
Here is what the author thought about it https://www mql5 com/en/forum/364798
"
Is an EA that collect Each pattern that occurs in the Market , open virtual orders with stop loss and take profit
EA will detect the best market pattern to open a real trade.
This is a deep learning expert advisor , EA will run for at least 3 to 5 months to start opening Real Trades.
"
"Ea will collect (x) candles to detect the patterns
Order will open in case a pattern repeat itself.
Code can be changed to make the ea detect
4 candles patterns,5,6,7 or 8
The more candles the more accuracy , less trades."
Some of the reviewers suggested:
"I'm thinking something like Run the EA in backtest collect the patters in CSV or serialzed objects, then when the EA is attached to real account and real time it will first start loading the files in memory to get the patterns and then it can start making real trades."
Could a coder here demonstrate is it possible to write the collected information data in a file, then loading it back into memory to continuously optimize and to use it for better trading.
It was proposed on but the author did not reacted on it. So I thought a coder in this thread could show us if it is feasible or the way how to do code it
Recently we have very aplaudable interesting situation. On MQL5 for first time a coder released soure file of self-adaptive deep learning strategy.
There is a variable that store all market patterns, in parallel a virtual trading system that open a buy and sell on each candle
Here is what the author thought about it https://www mql5 com/en/forum/364798
"
Is an EA that collect Each pattern that occurs in the Market , open virtual orders with stop loss and take profit
EA will detect the best market pattern to open a real trade.
This is a deep learning expert advisor , EA will run for at least 3 to 5 months to start opening Real Trades.
"
"Ea will collect (x) candles to detect the patterns
Order will open in case a pattern repeat itself.
Code can be changed to make the ea detect
4 candles patterns,5,6,7 or 8
The more candles the more accuracy , less trades."
Some of the reviewers suggested:
"I'm thinking something like Run the EA in backtest collect the patters in CSV or serialzed objects, then when the EA is attached to real account and real time it will first start loading the files in memory to get the patterns and then it can start making real trades."
Could a coder here demonstrate is it possible to write the collected information data in a file, then loading it back into memory to continuously optimize and to use it for better trading.
It was proposed on but the author did not reacted on it. So I thought a coder in this thread could show us if it is feasible or the way how to do code it
Attached File(s)
1
