Finding a set of indicators to feed into the neural network inputs. Discussion. A tool for evaluating the results. - page 10
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
... Fitting implies adjusting to the test suite, and here the finished system is just tested against it...
Works great on attractors :) I haven't tried it on pictures. I think the question is the structure of the row. I haven't pre-processed it.
If I find it, I'll send it to you. It's delphi, though...
Dabbled, but in a completely different application area. By the way, I failed to make non-linear PCA work. And linear one is too weak, I think.
In the extended space of Mercer's theorem, try to apply your data and solve the same problem, which is the same as the topic of association in neural networks discussed here.
Adjusting to a test set, or comparing to a test set - I don't see the difference. IMHO
There is a difference, and a big one (algorithmically and in results). But let us all stick to our opinions. If anyone is interested, please contact me in person.
If I find it, I'll send it to you. It's Delphi, though...
I'd appreciate it. I think the question was about pre-processing. I didn't do it, as I was just checking the network operation.
Greetings to all forum participants and visitors.
I want to offer all interested to discuss and search for a set of indicators best suited to feed into neural network inputs.
I can use profitability of neuronet in my program as an estimation tool, I can also lay out MQL4 Expert Advisor with trained neuronet. Within reasonable limits, of course.I have self-written (in java) perseptron with arbitrary number of layers and neurons in each layer I train it with genetic algorithm from JGAP library (http://jgap.sourceforge.net/).
The number of neurons in the first layer is equal to the number of inputs and in the second layer - arbitrarily, in the third layer 1 neuron. The neuronet produces trading signals (output of the neuronet >0.5 - buy, output of the neuronet<-0.5 - sell). The signals are processed by a self-written trading tester that based on the neuronet's signal reverses the position (or enters the market, if no position is opened). The target function of the genetic algorithm is the resulting profit. Such an approach, in my opinion, allows us to minimize all possible errors and bring the training as close as possible to the real trading. I export the trained network into MQL4 Expert Advisor and test it in the strategy tester of MT4. I form the inputs for the neural network in the MT4 indicator and download them into a file. The indicator and the Expert Advisor are formed by the program and are written to the files (less confusion and less errors).For me, 4-layer networks do not give more profit (usually less) than 3-layer ones, but they take longer to train.
I trained an 8-10-1 network for 4 days on a Core2 Quadro 2.3. 10 parallel threads with different initial populations competing to see "who has the most profit". 4 days passed about 4000 generations with 200 chromosomes in a population. The maximum profit was obtained in the first 2000 generations, beyond that, the profit did not increase. The biggest profit increase was in the first 100 generations.
I have checked results of this network in the MT4 strategy tester. I found out that the network almost never reaches the +-0.5 threshold and the trade signal is not triggered. The reason is unclear, i.e. I checked correctness of exporting to MQL4 (for the same values of inputs in Java and MQL4 the net gives the same values, maybe the whole incoming stream should have been submitted, not some random values). I lowered the threshold to 0.4 and it seems to work... Then I discovered that the Expert Advisor cannot reverse a position in one go... My Expert Advisor will close the bar and the price has time to move before the next bar. On the learning period (I am learning on 1-08-09 to 1-10-09) profit in MT4 was less than in my tester, on the testing period in MT4 (1-10-09 to 1-11-09), the net was profitable. I looked at what points of unprofitable entries occur and I got the impression that the data entering the neural network carries insufficient information...
I input neural network: (k=100)
I understand how indicators work, but I don't understand them and the market enough to choose minimal number of indicators by myself...I searched the forum and found (the author of post from which I took the idea, unfortunately, I don't remember):
studied in 10 threads, 200 generations, population size 200 for the period 1-08-09 to 1-10-09 (my java tester gives the results)network 9-10-1 : profit 10521
9-20-1 net : profit 10,434
9-30-1 network : profit 10361
9-50-1 network : profit 10059
result is good, but it seemed to be better with the previous version... i'll have to put it with previous inputs (i didn't save the results of last training)
Additional multipliers are needed to adjust values in the range from -1 to +1
Having read what I've written, I started thinking about a possible error in exporting the trained neural network to the Expert Advisor - it should be tested again.
P.S. Now I am writing a recurrent neural network of arbitrary structure (as I understand it, the recurrent neural network takes into account not only the value, but also the slope angle).
Among such brilliant neural network theorists, I would probably appear to be the black sheep. But someone has to make hamburgers and someone has to eat them. So I'll risk offering some interesting advice. You could use two indicators Force Index.mq4( period 3.53), DeMarker.mq4( period 3.53) for training neural network and connect changes in these indicators with normal time flow - time will be the third indicator. The indicators are simple and predict changes in the market with 100% reliability. In Force Index enter the level 0.0, and in DeMarker enter the levels 0.1 and 0.9. It is clear that in the code of these indicators we must change the dimension of periods from int to double.
Among such brilliant neural network theorists, I would probably appear to be the black sheep. But someone has to make hamburgers and someone has to eat them. So I'll risk offering some interesting advice. You could use two indicators Force Index.mq4( period 3.53), DeMarker.mq4( period 3.53) for training neural network and connect changes in these indicators with normal time flow - time will be the third indicator. The indicators are simple and predict changes in the market with 100% reliability. In Force Index enter level 0.0, and in DeMarker enter levels 0.1 and 0.9. It is clear, that in the code of these indicators we must change the dimension of periods from int to double.
so much time has passed... I wrote my genetic algorithm, made a recurrent forward propagation network with feedback from any subsequent layers to any previous ones... I haven't done the export of new neuristics into mql4... haven't got around to it... If I do, I'll keep your indicators in mind. :)