Noxa indicators for Neuroshell
- Thread starter val2004
- Start date
- Watchers 71
Hi arryex,
Thanks a lot for your message. I read all your study reports in this thread with big interest.
Do you compare your own SSA decomposition calculations with Noxa ones from the point of view of the repainting magnitude?
Frankly speaking, I believe there is some way to eliminate or at least diminish the SSA repainting - and the Noxa guys know it. What do you think about it?
Thanks a lot for your message. I read all your study reports in this thread with big interest.
Do you compare your own SSA decomposition calculations with Noxa ones from the point of view of the repainting magnitude?
Frankly speaking, I believe there is some way to eliminate or at least diminish the SSA repainting - and the Noxa guys know it. What do you think about it?
My previous articles shown that Noxa cycle is also reprinted for new data.
Noxa stated to use causality term related with his indicators, probably this related with unavailability of data. I am not sure this one is related with repainting issue.
To be frank, do you have Noxa or not? Noxa using lissajeous to define the parameter, trading strategy may not be found using NST optimization, most of time manual setting is applied.
I will post in next few days about impact new data on Noxa trading strategy.
Noxa stated to use causality term related with his indicators, probably this related with unavailability of data. I am not sure this one is related with repainting issue.
To be frank, do you have Noxa or not? Noxa using lissajeous to define the parameter, trading strategy may not be found using NST optimization, most of time manual setting is applied.
I will post in next few days about impact new data on Noxa trading strategy.
donkingdon
Newbie
- Messages
- 9
- Likes
- 0
Interesting discussion.
I can tell, CSSA doesnt repaint. For that set the training range to other than the value by default. Set it to less than the number of bars in your chart therefore there is no new training of the indicator when you receive a new bar. Try it. You will see it does not repaint.
Oh I see. maybe you meant The Noxa SSA? this one you can get free on their website? This one repaints because it is not causal. But their Noxa CSSA does not repaint. I don't know how they do it.
donkingdon
Newbie
- Messages
- 9
- Likes
- 0
very nice ..... if I apply my permission ... trading in my work .... and I would use ....
indonesianya mana yah.... ?
indonesianya mana yah.... ?
Asli orang Sunda (Bandung) tinggal di Bpn kaltim:cheesy:
donkingdon
Newbie
- Messages
- 9
- Likes
- 0
I am very happy to be friends with you .... originating from one country .... get on with it ... I asked ... if until now you still ... using n.n if I may more closely ... This addresses my emai: *************** me so interested in your strategy, I live in central java-cilacap
Last edited:
Hi,
I just read the explanation how Noxa build the causal on SSA.
Check on the help: There are crude ways to make SSA cycles causal:
· the easiest way would be to lag the cycle by m-histories,
· Another way would be to "re-run" the entire SSA calculation in a rolling window, adding one bar each time (end-point version).
I did the similar steps..see here the result on my own SSA trend. All the values within train bars are kept. Now My SSA will not change after new bars coming:smart:
I just read the explanation how Noxa build the causal on SSA.
Check on the help: There are crude ways to make SSA cycles causal:
· the easiest way would be to lag the cycle by m-histories,
· Another way would be to "re-run" the entire SSA calculation in a rolling window, adding one bar each time (end-point version).
I did the similar steps..see here the result on my own SSA trend. All the values within train bars are kept. Now My SSA will not change after new bars coming:smart:
Attachments
But they wrote bellow
"However we found both methods very limited in their predictive skills; they offered little guidance for trading so we devised a proprietary causal version for SSA that is particularly well-suited for use in trading strategies."
I think to make prediction, I prefer to use NN instead of CSSA..see this thread if you interested
http://www.trade2win.com/boards/tra...l-network-indicator-mt4-using-neuroshell.html
Noxa indicator is not adaptive, we may need to reset the parameter when the market characteristic is changed. There is no guarantee for a single parameter setting of indicators will provide a profitable trading strategy during OOS, and again you set them manually.
I have some difficulties to use NST optimization to give the best noxa parameters.
One benefit for the user is, if the market still follow the last cycle..if change then you need to redefine the cycle. That is why try using matlab (if successful:cheesy
http://www.trade2win.com/boards/tra...l-network-indicator-mt4-using-neuroshell.html
Noxa indicator is not adaptive, we may need to reset the parameter when the market characteristic is changed. There is no guarantee for a single parameter setting of indicators will provide a profitable trading strategy during OOS, and again you set them manually.
I have some difficulties to use NST optimization to give the best noxa parameters.
One benefit for the user is, if the market still follow the last cycle..if change then you need to redefine the cycle. That is why try using matlab (if successful:cheesy
But they wrote bellow
"However we found both methods very limited in their predictive skills; they offered little guidance for trading so we devised a proprietary causal version for SSA that is particularly well-suited for use in trading strategies."
Hi Phepi,
I have following comments about your template:
- The trading strategy is made based on the prediction, but I found that the range for optimization are overlap (see the system equity between prediction and trading strategy). Not sure whether it will impact the optimization or not, but i should avoid it.
-The input prediction is quite complex, probably the optimization will take times.
- Check the prediction result and actual ..you can see whether the prediction can predict the actual?
Theoretically, I prefer to use the simple one, ensure the prediction is good then used it on trading strategy.
Someone else may give the feedback...
I have following comments about your template:
- The trading strategy is made based on the prediction, but I found that the range for optimization are overlap (see the system equity between prediction and trading strategy). Not sure whether it will impact the optimization or not, but i should avoid it.
-The input prediction is quite complex, probably the optimization will take times.
- Check the prediction result and actual ..you can see whether the prediction can predict the actual?
Theoretically, I prefer to use the simple one, ensure the prediction is good then used it on trading strategy.
Someone else may give the feedback...
Attachments
amarozo
Newbie
- Messages
- 8
- Likes
- 0
Hello Arry, thank you very much for great analyse , tomorow I will post new one or will change little bit. So far I have find out, that Noxa CSSA is very well working with clustering indicators and FFT. (yes, I give to this trading strategy, from prediction some trailing stops)
Phepi
, tomorow I will post new one or will change little bit. So far I have find out, that Noxa CSSA is very well working with clustering indicators and FFT. (yes, I give to this trading strategy, from prediction some trailing stops)Phepi
Hi Phepi,
This is small correlation analysis between predicted and actual from your prediction strategy.
The calculated correlation is quite small and R2 is negative, means your prediction value may going in different direction with your actual data. Neuroshell trader do not provide this values, hence we can check only the prediction strategy result. You may found it from Neuroshell2 to evaluate prediction quality.
This is small correlation analysis between predicted and actual from your prediction strategy.
The calculated correlation is quite small and R2 is negative, means your prediction value may going in different direction with your actual data. Neuroshell trader do not provide this values, hence we can check only the prediction strategy result. You may found it from Neuroshell2 to evaluate prediction quality.
amarozo
Newbie
- Messages
- 8
- Likes
- 0
Hello all I see a nice forum is here, I am also little playing with Noxa CSSA 1.3 (just have buy this addon on 5. october, just learn how to use it) and have made a little profitable template, here do take a look, want share with you all btw any tips how can be even more efective?
thanks Amarozo
P.S. First I have extract right cycle for CSSA, I have re-run SSA and get best output for setting Noxa CSSA, then I have combined this results with clustering indicators on close and lets deside NN how close it is to sell or to buy. After that i have give this results to Tprop4 basic neural network algorithm to choose the right one
thanks Amarozo
P.S. First I have extract right cycle for CSSA, I have re-run SSA and get best output for setting Noxa CSSA, then I have combined this results with clustering indicators on close and lets deside NN how close it is to sell or to buy. After that i have give this results to Tprop4 basic neural network algorithm to choose the right one
