Machine learning for robots - page 9
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I made an experiment to train the Expert Advisor on USD-Frank and tried to compare its tests on other pairs containing basic or quoted currency.
The settings in all tests were identical and this is what came out:
The result shows that both currencies introduce their own patterns in the pair and the neuronet finds them.
I do not understand why the trend is not confirmed only on the euro-franc.
If you want to run the test, I can send you the Expert Advisor.
I made an experiment to train the Expert Advisor on the dollar-franc pair and tried to compare its tests on other pairs containing base or quoted currencies.
The settings in all tests were identical and this is what came out:
USDCHF
USDCAD
USDJPY
GBPCHF
CADCHF
AUDCHF
EURCHF
The result shows that both currencies introduce their own patterns in the pair and the neuronet finds them.
It is not clear why the trend is not confirmed only on EURCHF, can someone explain.
If you want to run the test yourself, I can drop the advisor in my personal account.
Well, the test please on EURGBP, and AUDNZD - these are almost analogues of EURCHF
And a test please on EURGBP, and AUDNZD - these are almost analogous to EURCHF
EURGBP
AUDNZD
Thanks! )))
EURGBP
AUDNZD
And for which period (__ to __) and on which timeframe, if it is not a secret, was the result obtained?
And over what period (from __ to __) and on what timeframe, if it is not a secret, is the result obtained?
I made an experiment in training the Expert Advisor on the dollar-franc pair and tried to compare it with other pairs containing basic or quoted currency.
The settings in all tests were identical and this is what came out:
USDCHF
USDCAD
USDJPY
GBPCHF
CADCHF
AUDCHF
EURCHF
The result shows that both currencies introduce their own patterns in the pair and the neuronet finds them.
It is not clear why the trend was not confirmed only on EURCHF, can someone explain.
If you want to run the test yourself, I can drop the advisor in my personal account.
Interesting to see GBPJPY, EURJPY, AUDJPY. Somehow it seems to me that one of these pairs would be better.
Is this EA trend-following or flat?
I want to disappoint you, in compiled form you won't be allowed here. Yes, there is no such ban in the forum rules, but the thread will be taken down. I recently had a similar experiment taken down in the English part. Check my personal account and I'll post recommendations on where to do it.
In English-speaking forum some fairly harmless things are torn down that are easily discussed in Russian. The nuts are tightened tighter there than here. And this is a fact noticed by many.
2008-2018 H4 MetaQuotes-Demo
Results of the total test run of the currency pairs usdchf and eurusd for one working week from 24.09 to 28.09 (with spread 10 and current spread).
usdchf h4 h1 m5
10 69 -1336 70 -1414 70 -14
text 70 -2349 71 -2428 72 -2536
eurusd h4 h1 m5
10 145 -572 146 -647 150 -870
text 142 -1680 145 -1672 147 -1822
The only consolation is that the number of trades is quite sufficient, roughly comparable to the number of working trends. But the processing of the trends themselves is completely beside the point!
Results of the total test run of the currency pairs usdchf and eurusd for one working week from 24.09 to 28.09 (with spread 10 and current spread).
usdchf h4 h1 m5
10 69 -1336 70 -1414 70 -14
text 70 -2349 71 -2428 72 -2536
eurusd h4 h1 m5
10 145 -572 146 -647 150 -870
text 142 -1680 145 -1672 147 -1822
The only consolation is that the number of trades is quite sufficient, roughly comparable to the number of working trends. But the processing of the trends themselves is completely out of order!
Perhaps you have not run the tests on MetaQuotes-Demo, but, in addition, have used other timeframes.
I trained only on OHLC USDCHF H4 - it's an experiment for MetaQuotes-Demo because there is a large base, the quotes from other providers may be significantly different, and even GMT.
The problem of unification of learning, so the Expert Advisor would not be sensitive to differences in quotes and that it summarizes information from different timeframes, is another problem.
For this I experiment with formulas of defining training patterns and integral price bar characteristics.
Recently an interesting solution was suggested by one of the programmers in the English part of the forum, if you have any ideas in this direction, please let me know.
https://www.mql5.com/en/forum/281402/page4