Counsellors to whom. Lots of them and for free! - page 7
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
Thank you very much for your oscillator, it is very informative indeed!
Yes, zfs, the oscillator is interesting! Except that it needs to be inverted. Blue on the plus side, red on the minus.
It would be perceived better visually, imho. And the signals would be more logical
Probably more logical.
Yes, zfs, the oscillator is interesting! Except that it needs to be inverted. Blue on the plus side, red on the minus.
It would be perceived better visually, imho. And the signals would be more logical
probably makes more sense.
I'm already used to it that way. Especially since the oscillator moves behind the price - there is no point in flipping. You just have to make sense of it as it is. But I can say that at first I had the same idea.
I'll see what the problem is. Maybe there are some input parameters d* missing in the settings?
I doubt it, the point is that it is not in a single instance.
So far I've found 2 types of oscillator interpretation (apparently there are only 2):
1. on breakdown, the signal appears on breakdown, i.e. if there is no breakdown, you can play on breakdown at the indicator boundary values (found on the pound 5 minutes).On the trend inside the channel with a stop.
2.The oscillator follows the price and reflects all the troughs and tops. The more value of the red bar, the more profitable to buy, and the more profitable to sell. Unlike the previous option.
I doubt it, the point is that it is not in a single instance.
So far I have found 2 types of oscillator interpretation (apparently there are only 2):
1. on the breakdown, the signal appears on the breakdown, that is, if there is no breakdown, you can play on the rebound at the boundary values of the indicator (found on the pound, 5 minutes), in the trend inside the channel with a stop.
2.The oscillator follows the price and reflects all the troughs and tops. The more value of the red bar, the more profitable to buy, and the more profitable to sell. In contrast to the previous option.
SSB essentially uses neural network technology, as each trading signal from an oscillator or an indicator is assigned a weight. The difference is that the weights have only three discrete levels -1, 0, 1.
If you had a better tester, you could have sampled more capacity. Or maybe you don't need to do it at all, as the retrained network is just a fake?
This is the first time I've traded on the minutes and even in profit. I used to avoid small timeframes because I thought they were too noisy. But it turned out that the noise is not so noisy and periods of transitions from one stationary state to another are just directly proportional to timeframe periods.
If I had a better tester I could make sampling more capacious. >> Or maybe not at all, since overfitted network is a null fit?
Worth a try in my opinion!
For the first time I`ve traded on forex on few minutes and i`ve earned profit. I used to avoid small timeframes, because I believe they are too noisy. But it turned out that the noise is not so noisy and periods of change from one stationary state to another is just directly proportional to timeframe periods.
I also get very interesting results with the minutes, oddly enough. It's true that I use only my own indicators. I have not registered proportions with the higher prices yet, I have not analysed them. I have not yet analyzed it. Can you explain or show me how it may occur?
SSB essentially uses neural network technology as each trading signal from an oscillator or indicator is given a weight. The difference is that the weights have only three discrete levels -1, 0, 1.
If the tester were more powerful, the sampling could be more capacious. Or maybe I don't need it at all, since the retrained network is a naked fit?
This is the first time I've traded on the minutes and even in profit. I used to avoid small timeframes because I thought they were too noisy. But it appeared the noise is not so big and periods of transitions from one stationary state to another are just proportional to timeframe periods.
I also started with large timeframes. The idea is definitely a good one. I did something similar at Wealth Lab, also a kind of neural network. (for stocks). Finam now allows me to trade stock derivatives in Metatrader - I think it is even more relevant there. In general, I think we should expand the number of instruments - create new ones. Principle -1,0,1 should not be widened, in principle you already use 2 factors on some oscillators, you can widen the number of factors. A package of new oscilloscopes to provide with the program. I think the Bollinger lines should appear, I suggest to use the ideas of Miroslav: Oscillator of ATR, Bulls Bears Power (published on forum); Donchan, Keltner channels, MFI etc. In addition I propose to limit the number of selected factors as with the increase of factors we lose reliability - after all the system should be simple and on a 5 minute chart should do several transactions a day, and not 1. Therefore I recommend to limit the number of selected factors, while increasing the number of their variants.
The idea is definitely a good one. I did something similar at Wealth Lab, also a kind of neural network. (for stocks).
Which idea do you mean (Reshetov's)?
Principle -1,0,1 is not worth expanding in principle you already use 2 factors for some oscillations, you can expand the number of factors. I think that Bollinger lines should appear. Donchian channels, Keltner...
And sometimes three indicators give better results than 10 or more. Only you have to use them more... um... practical or something. :) That's why I'm in favour of higher sampling. But I'm also in favour of channels and Bollinger.
Which idea are you referring to (Reshetov's)?
And with me sometimes three indicators give better results than 10 or more. Only you have to use them more... um... practical or something. :) That's why I'm in favour of higher sampling. But I'm also in favour of channels and Bollinger.
>> Right.
What's the point of upsampling? I don't see the point. Everything is relative. Well, if you want some indicator to exceed some level, you can introduce an additional factor. And the level itself can be optimized as a period.
What's the point of upsampling? I don't see the point. Everything is relative. Well, if you want some indicator to exceed some level, you can introduce an additional factor. And the level itself can be optimized as a period.
Yes, you can do it as a factor, that's what I do. But I would like to do it as additional discretization :)