How do you work with neural networks? - page 5
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
How does GA relate to NS and regression?
NS is a method.
GA is a method.
"Use GA instead of NS" sounds crazy. It's like "replace the heart with an exhaust gas analyzer."
I'm sorry. (chuckles)
One problem with neural networks, as well as with other TCs that do not use neural networks - a neural network will always find a pattern at any given time interval (training or optimization section), then there is the same question - will this pattern work (bring profit) in the future?
The answer is banal: If the patterns found in the past do not contradict themselves in the future, there will be profit.
For example, if in the past, on which the net was trained, the sideways trend prevailed, and in the future a prolonged up trend or a down trend began, we can hardly expect profit, because the net will be trained for the rebound of a strong price movement in one direction. But if the previous uprend turns into a downrend or vice versa in the future, a normal grid should give profits.
For example, if in the past, on which the grid was trained, the sideways trend prevailed, and in the future a prolonged up trend or a down trend began, profit is unlikely to occur, because the grid will be trained for the rebound of a strong price movement in one direction. But if the previous uprend turns into a downrend or vice versa in the future, a normal grid should give profits.
This concerns not only NS, but also other systems
That there is a "yes" answer - there are such patterns ))))
joo : Will traders stop trading?
No, because the thirst for profit is indestructible in man ))))
joo : Zoo: Will traders buy reliable information confirming that the answer is "No"? Or would they prefer not to know the answer to that question? (rhetoric if anything).
They will "fight" to find the answer to that question and get a "yes" answer )))
The solution of this issue is simple - we should teach the net on the time interval when all types of motion are present. It can be sideways, upwards or downwards. Of course, we must understand that if the net is trained only on the up trend, it will fail on the down trend )))).
Of course, it should be understood that if the net is only trained on up-trend, it will lose on down-trend ))))
A well-trained grid in such circumstances should not fail. I.e. the sign of a well-trained grid is at least a fitting on the history chart followed by a positive "profit" on an inverted chart as an additional OOS.
If the grid will lose on the inverted chart, then it is much worse than any primitive TS set up either only on the trend or only on the counter-trend.
Try Matlab ANFIS with Candlestick patterns (svechnymi combinacijami) as described here'
Interesno kakije budut'taty?
:-)
Valera
Try Matlab ANFIS with Candlestick patterns (svechnymi combinacijami) as described here'
Interesno kakije budut'taty?
:-)
Valera
A well-trained grid in such circumstances should not lose money. I.e. the sign of a well-trained grid is at least a fit on the history chart followed by a positive "profit" on an inverted chart as an additional OOS.
If the grid will lose on the inverted chart, then it is much worse than any primitive TS, adjusted either only for the trend or only for the counter-trend.
To be honest, I don't really understand your point.