Trading Strategies Based On Digital Filters - page 44
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
digital filters
Hi Richcap,
I just finished evaluating NOXA indicators on TRADE2WIN (Krzysiaczek99) and looking again to this thread with a new expirience on CSSA.
If you designed all this stuff as a stand alone application than you are a God programmer. How long it took you ??
I joined this thread very late when suddenly it stopped. In my opinion this method was never evaluated properly, now I'm using Neurshell with MT4 and it is very easy to evaluate this method with trading strategy and also use filters as an addition to another strategies. This is my plan for next weeks, I'm also going to implement methods from Codebreaker forexfactory 'optimized trend trading' thread using NS as in my opinion this is the most effective way to work.
MESA is maybe not the best idea, would be good to evaluate usage of GOERTZEL + advanced Fourier indicators, I believe that the proper combination + noise filtering will give good results.
Krzysztof
noise definition
I think in Codebreaker thread 'noise' is discussed quite deeply, than you can cross check with your definitions.
Krzysztof
Hi Richcap,
I just finished evaluating NOXA indicators on TRADE2WIN (Krzysiaczek99) and looking again to this thread with a new expirience on CSSA.
If you designed all this stuff as a stand alone application than you are a God programmer. How long it took you ??
I joined this thread very late when suddenly it stopped. In my opinion this method was never evaluated properly, now I'm using Neurshell with MT4 and it is very easy to evaluate this method with trading strategy and also use filters as an addition to another strategies. This is my plan for next weeks, I'm also going to implement methods from Codebreaker forexfactory 'optimized trend trading' thread using NS as in my opinion this is the most effective way to work.
MESA is maybe not the best idea, would be good to evaluate usage of GOERTZEL + advanced Fourier indicators, I believe that the proper combination + noise filtering will give good results.
KrzysztofHi Krzysztof,
well, it took some time to port available "C" code to metatrader. I was lucky enough to find a lot of good code. You know that mql is a subset of 'C', so the porting is straightforward.
I have followed some very interesting thread about neuroshell + metatrader in the past, but I must confess that I never tried because I know that NS+metatrader is a pain in terms of reliability and I don't want to waste time after "dll hell" or similar problems. If I'm going to evaluate neural networks I will do it my way, as I did for digital filtering and spectrum analiys.
I would like to attach hereafter my digital filter indicator, as a little gift back to whom enriched this very stimulating 3d with his/her patience, but I'm not allowed yet
Good night.
How do we choose the parameters for the CL Cycle indicators?
How do we choose the parameters for the CL Cycle indicators?
Dvarrin,
Which Cycle indicators are you talking about? It's been a long while since I have posted anything, so my mind is drawing a blank.
Also, I have spent considerable time with VK's DF method in the past. I believe he has 8 different entry signals for long trade setups and 8 for shorts. There is a lot of information there, so fully optimizing the strategy to trade automatically would be VERY difficult IMHO. I believe the creators also found this road to not be the optimum one.
The key to EVERYTHING is to obviously find the optimizum P/D values. If there are questions about the Spectrum Analyzer and such please feel free to ask and I will give my $.02.
Best,
cl
Hi all,
First I developed a digital filter, using known algorythms and open source 'C' code. So I it is not needed offline filter coefficent generation any more.
Then I developed a MESA spectrum analyser to extract significant frequencies and in the end I developed adaptive FATL-SATL, FTLM,STLM , PCCI and RBCI indicators for metatrader.
My suspect is that spectrum analisys failures, together with the ever changing nature of markets and some bad assumption about what "noise" is in maket price variations tend to invalidate this approach.
1. I believe there is an external Digital Filter Indicator that allows the creation of filters on the fly so you don't have to use the DFG.
2. Did you develop ANOTHER Mesa Spectrum Analyzer in addition to the one already available? What do you mean by adaptive?
3. IMHO, you have create an "edge". There are many approaches out there to isolate dominant cycles. Doing this with a high degree of accuracy is the key. The DFG is Garbage In - Garbage Out. It is soley dependant on the paramaters you pass to it.
Best,
cl
Dvarrin,
Which Cycle indicators are you talking about? It's been a long while since I have posted anything, so my mind is drawing a blank.
Also, I have spent considerable time with VK's DF method in the past. I believe he has 8 different entry signals for long trade setups and 8 for shorts. There is a lot of information there, so fully optimizing the strategy to trade automatically would be VERY difficult IMHO. I believe the creators also found this road to not be the optimum one.
The key to EVERYTHING is to obviously find the optimizum P/D values. If there are questions about the Spectrum Analyzer and such please feel free to ask and I will give my $.02.
Best,
clHi Clahn04,
I mean the cycles as shown on post 273 and 275? Are they RBCI? how do we choose the parameters for P1, D1, P2 and D2?
Hi Clahn04, I mean the cycles as shown on post 273 and 275? Are they RBCI? how do we choose the parameters for P1, D1, P2 and D2?
RBCI is a bandpass filter. It's equation is FATL(k) - SATL(k). This is very different than the cycles you see there. The cycles in the posts are bandpass filters, but they are made from the DFG software.
The posts around there clearly show 2 different ways of calculating the cycles. I would highly reccomend rereading them. To summarize, after finding the dominant peaks with the spectrum analyzer, you can create the cycles in a couple of different ways.
1. Create a cycle in which you isolate a dominant peak (i.e., since we are using bandpass filters, you would want to "pass" that peak and filter all the rest.
2. Isolate several dominant peaks and create a more robust cycle. This version will not have a "sinewave" look to it.
Read around post #253. ;-)
cl
RBCI is a bandpass filter. It's equation is FATL(k) - SATL(k). This is very different than the cycles you see there. The cycles in the posts are bandpass filters, but they are made from the DFG software.
The posts around there clearly show 2 different ways of calculating the cycles. I would highly reccomend rereading them. To summarize, after finding the dominant peaks with the spectrum analyzer, you can create the cycles in a couple of different ways.
1. Create a cycle in which you isolate a dominant peak (i.e., since we are using bandpass filters, you would want to "pass" that peak and filter all the rest.
2. Isolate several dominant peaks and create a more robust cycle. This version will not have a "sinewave" look to it.
Read around post #253. ;-)
clThank you Clahn04. I'll read again from post 253 on. I also saw your CL_slope and CL_slope 2 indicators at the Simba Con Man Thread. How do you build them? Have you found a successful strategy using them?
Thank you Clahn04. I'll read again from post 253 on. I also saw your CL_slope and CL_slope 2 indicators at the Simba Con Man Thread. How do you build them? Have you found a successful strategy using them?
They were simple indicators used to show the changes in Centered MA's (Snakes). This method is taken straight from JM Hurst's books (well, the idea is anyways....Hurst uses displaced MA's). As a stand alone, Snakes are very hard to use. They require you to know the dominant cycle. IF you do, then you can calculate the half cycles very easily. This is true with many indicators. Many people use CCI 50 or 100 crossing the zero line as an example because it passes the "looks" test. For me, the "looks" test isn't that imortant as the "why" test. You can simply set the CCI to the half cycle of the dominant cycle for a given security/currency and it will tell you much much more than an arbitrary setting (this is just an example...i do not use CCI in my trading).
All of my methods are based on two things: Cycles and Noise(how to reduce it). Once you have a stable base to deal with both of these the fleshing out of a system becomes much more attainable.
ATCF took this approach. I spent many many hours studying the method. Their "stable base" is the interplay of SATL/STLM. These 2 indicators are used to give the current "state" of the market. FTLM, RBCI, PCCI, FATL, etc, are all used to model the intricate swings that longer term cycles miss. RBCI/PCCI are used as volatility measures for exits.
Best,
cl