Explanation Choose the Best DNN
explanation Choose the Best DNN
before using "Choose The Best" you must first make the following settings.
in this mode EAs can select the DNN Model used based on the Month, take a look at the following image:
in the picture above "SetBestDNN" has 12 numbers (12 months) and each number is 3,8,9,1,3,6,7,2,20,14,16,18, which means that in January EA uses Model 3, in February uses model 8, March model 9, April model 1, and so on. You can use a minimum number of 1,1,1,1,1,1,1,1,1,1,1 and a maximum of 30,30,30,30,30,30,30,30,30,30,30,30 you can use random numbers to find the best strategy.
How to find the best strategy with this feature?...
1. you must have Historical data for the pair you want to backtest for at least 1 year. if not then please download it at Tools> Historical Senter
3. Set the desired Close position, you can choose Trailing or other Close settings according to the desired profit target, and also consider realistic profit targets.
4. try one by one starting from 1,1,1,1,1,1,1,1,1,1,1,1,1, this means that for 12 months you will use Model 1, and if the results are unsatisfactory or MC, then what you have to do is change the DNN model in the MC month, for example if the MC occurs in March then you have to change the DNN for example 1, 1,2,1,1,1,1,1,1,1,1,1,1, this means that in March EA will automatically change to model 2, and in April change to model 1 again, do this in the month of MC or the results are not satisfactory, until you find a good strategy in one year.
5. If you find a strategy that can last for 1 year, I recommend testing it in the previous 12 months.
If it passes 1 year = Good Strategy
If it passes 2 years = Very good strategy
If it passes more than 3 years = Excellent Strategy
The point is that the more data that is tested and the results are good, the more confident a strategy is to be used for Forward Test.
The following is a map of the creation of the DNN Strategy by Choose The Best
Notes:
If you don't find a good strategy, try to change the TimeFrame or Close Position or Open Position, because this is very related. If SetBestDNN = 3,8,9,1,3,6,7,2,20,14,16,18 and tested on 5m TimeFrame and produces a less good strategy, it does not mean that on TimeFrame 1H the strategy is bad too, as well as Open position and Close position. if SetBestDNN results are bad on Close_By_Amount, it might be good with Close_By_Percent.
Komentar
Posting Komentar