Better NN EA development - page 26
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
SVM_v_27a_EURUSD
//////////////////////////////////////////
#define NR_BARS1 6000
#define NR_BARS2 2000
extern string Symb0 ="EURUSDm";
extern double Lots =0.01;
extern int diff1 =true;
extern int decimalplace =14;
//extern int svmtype =0; //0-libsvm,1-svmlight
extern int BasePeriod =PERIOD_M5;
extern int TrainPeriod =PERIOD_H1;
extern int class_type =10; //0-multiclass,
//1-class(0,1,-1),
//2,3-class(-3,-2,-1,0,1,2,3)
//4-class(-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6)
extern int TrainStopLoss_buy=20;
extern int TrainStopLoss_sell=20;
extern int TrainTakeProfit_buy=20;
extern int TrainTakeProfit_sell=20;
extern int x1= 120;
extern int x2= 120;
extern int TradeStopLoss =60; //30-50pips
extern int TradeTakeProfit =60;
extern int FutureWindow =120;
extern int Signal =8;
extern bool UseRamDisk =true;
extern datetime BeginDate = D'2008.04.02 08:15';//D'2008.05.01 14:00';//D'2008.04.02 08:15';
extern bool ReverseSignal=false;
extern bool TradeEnabled =false;
//////////////////////////////////////////////////
5_bar_SVM_input
Could someone outline the procedure to employ "5_bar_SVM_input" for the 5_bar indicators posted previously?
Stop trying to teach an NN to predict. Instead, teach it how to trade. There is a difference.
Very good, do you have any idea how to teach an ea to be idle? You have no result (profit/loss) if ea is idle in market range, so how to generate the teaching input? (This is a serious question, not ironic at all!)
Regards
bbop
teaching input
i think the input compined with a trading rule is the pattern of the last bars (dependent of TF, or multi TF NN input). but normalized to get unit vectors of the price difference. so it is equal if the price is 1.5445 or 107.40 or 0.7665 and so on.
for learning i would open random orders. memorize every NN weights to every order and have a basic standard exit rule for the beginning. if the trade was give positive results this is the output what the NN should give (reward. reinforcement learning) and if the trade was a loss then adjust the weights of the NN.
and then combine this with a exit reward NN (maybe this will work). you could also use SOM for classification instead patterns and analyze what happend after a memorized SOM. this is similar what KLOT did in the russian forum. important is not the forming of a pattern. important is what pattern comes after a ready pattern. this could be analyzed with data mining. for example.
you have 11 classes of a SOM. what class forms after the class number 6? if you find out that after every class number 6 the class number 2 will develop to 80% and you know what price action the class no. 2 give, then you could take the right order.
for idle state it depents on the trading rule. if you for example use suppport resistance it will be idle between them and go for thinking if it tests the s or r twice, three times etc.
this is my opinion. if it works? don't know. i will test it
PriNova
SVM_v_27a_OIL
//////////////////////////////////////////
#define NR_BARS1 6000
#define NR_BARS2 2000
extern string Symb0 ="OIL";
extern double Lots =0.1;
extern int diff1 =true;
extern int decimalplace =14;
//extern int svmtype =0; //0-libsvm,1-svmlight
extern int BasePeriod =PERIOD_M5;
extern int TrainPeriod =PERIOD_H1;
extern int class_type =10; //0-multiclass,
//1-class(0,1,-1),
//2,3-class(-3,-2,-1,0,1,2,3)
//4-class(-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6)
extern int TrainStopLoss_buy=50; //15-100 pips H1
extern int TrainStopLoss_sell=50;
extern int TrainTakeProfit_buy=50;
extern int TrainTakeProfit_sell=50;
extern int x1= 120;
extern int x2= 120;
extern int TradeStopLoss =100;
extern int TradeTakeProfit =100;
extern int FutureWindow =120;
extern int Signal =8;
extern bool UseRamDisk =true;
extern datetime BeginDate = D'2008.04.02 08:15';//D'2008.05.01 14:00';//D'2008.04.02 08:15';
extern bool ReverseSignal=false;
extern bool TradeEnabled =false;
//////////////////////////////////////////////////
//////////////////////////////////////////
SVM_v_27a_EURUSD
//////////////////////////////////////////
#define NR_BARS1 6000
#define NR_BARS2 2000
extern string Symb0 ="EURUSDm";
extern double Lots =0.01;
extern int diff1 =true;
extern int decimalplace =14;
//extern int svmtype =0; //0-libsvm,1-svmlight
extern int BasePeriod =PERIOD_M5;
extern int TrainPeriod =PERIOD_H1;
extern int class_type =10; //0-multiclass,
//1-class(0,1,-1),
//2,3-class(-3,-2,-1,0,1,2,3)
//4-class(-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6)
extern int TrainStopLoss_buy=20;
extern int TrainStopLoss_sell=20;
extern int TrainTakeProfit_buy=20;
extern int TrainTakeProfit_sell=20;
extern int x1= 120;
extern int x2= 120;
extern int TradeStopLoss =60; //30-50pips
extern int TradeTakeProfit =60;
extern int FutureWindow =120;
extern int Signal =8;
extern bool UseRamDisk =true;
extern datetime BeginDate = D'2008.04.02 08:15';//D'2008.05.01 14:00';//D'2008.04.02 08:15';
extern bool ReverseSignal=false;
extern bool TradeEnabled =false;
//////////////////////////////////////////////////
i think the input compined with a trading rule is the pattern of the last bars (dependent of TF, or multi TF NN input). but normalized to get unit vectors of the price difference. so it is equal if the price is 1.5445 or 107.40 or 0.7665 and so on.
for learning i would open random orders. memorize every NN weights to every order and have a basic standard exit rule for the beginning. if the trade was give positive results this is the output what the NN should give (reward. reinforcement learning) and if the trade was a loss then adjust the weights of the NN.
and then combine this with a exit reward NN (maybe this will work). you could also use SOM for classification instead patterns and analyze what happend after a memorized SOM. this is similar what KLOT did in the russian forum. important is not the forming of a pattern. important is what pattern comes after a ready pattern. this could be analyzed with data mining. for example.
you have 11 classes of a SOM. what class forms after the class number 6? if you find out that after every class number 6 the class number 2 will develop to 80% and you know what price action the class no. 2 give, then you could take the right order.
for idle state it depents on the trading rule. if you for example use suppport resistance it will be idle between them and go for thinking if it tests the s or r twice, three times etc.
this is my opinion. if it works? don't know. i will test it
PriNovaI am looking for anyone who have a working Cortex Neural Network Software ( i think the last known version was version 5). If you have a working application please let me know. i need it very urgently to test and train some EA in metatrader 4 . I did purchase a copy from Neural Networks Tutorials and Software but it seems to be missing a "wait.dll" file. They also have a free copy but both of them don't seem to work since they are missing the "wait.dll"
Anyone who has used or have experience with this software . I am basically using it to train and test my Metatrader 4 EAs
I am running it on Windows XP professional with service pack 2.
Any help or suggestion is well appreciated. i need to Finish the EA by this weekend
Hello everyone,
It looks like that probably a person have imitate Better's ea successful.
Here is a list of his ea's orders.
Let us look at it,may be instructive.
Hello everyone,
It looks like that probably a person have imitate Better's ea successful.
Here is a list of his ea's orders.
Let us look at it,may be instructive.Where did you find this?
Bogie NN-8
Where did you find this?
There is no mystery, this is the no less famous William Boatright (the second on ATC 2007) and your Bogie NN-8. You will find it here:
Bogie Enterprises
Also, you can see the forward-test here:
forex blog by ckowyong - finding the best forex expert advisor…
That's it.