How can I predict future values of currency prices, in a neural network time series trained with daily data, predicting 30 days into the future from the last day's data.

I am using neural network time series tool to predict the values of currency exchange rates. I have trained the network using NARX model, delay - 2 hidden layers - 2 I trained the network with 4000 timesteps of data. I have also trained another network with hourly data - around 70000 timesteps data.

I used NaN values in the inputs and the targets, for the future 30 days that the network has to predict.

I am getting values like 0.0323, 0.0339, etc., close to zero and nowhere closer to the values that i need to see, somewhere around 0.9343 something.

I opened the loop and then also it came like that. I closed the loop, and increased the delay to 21, and then also it came like that.

I will paste my code here.

| | *

% Solve an Autoregression Problem with External Input with a NARX Neural Network
% Script generated by NTSTOOL
% Created Sat Sep 21 13:26:03 IST 2013
%
% This script assumes these variables are defined:
%
%   input - input time series.
%   output - feedback time series.

inputSeries = tonndata(input,false,false);
targetSeries = tonndata(output,false,false);

% Create a Nonlinear Autoregressive Network with External Input
inputDelays = 1:2;
feedbackDelays = 1:2;
hiddenLayerSize = 2;
net = narxnet(inputDelays,feedbackDelays,hiddenLayerSize);

% Choose Input and Feedback Pre/Post-Processing Functions
% Settings for feedback input are automatically applied to feedback output
% For a list of all processing functions type: help nnprocess
% Customize input parameters at: net.inputs{i}.processParam
% Customize output parameters at: net.outputs{i}.processParam
net.inputs{1}.processFcns = {'removeconstantrows','mapminmax'};
net.inputs{2}.processFcns = {'removeconstantrows','mapminmax'};

% Prepare the Data for Training and Simulation
% The function PREPARETS prepares timeseries data for a particular network,
% shifting time by the minimum amount to fill input states and layer states.
% Using PREPARETS allows you to keep your original time series data unchanged, while
% easily customizing it for networks with differing numbers of delays, with
% open loop or closed loop feedback modes.
[inputs,inputStates,layerStates,targets] = preparets(net,inputSeries,{},targetSeries);

% Setup Division of Data for Training, Validation, Testing
% The function DIVIDERAND randomly assigns target values to training,
% validation and test sets during training.
% For a list of all data division functions type: help nndivide
net.divideFcn = 'dividerand';  % Divide data randomly
% The property DIVIDEMODE set to TIMESTEP means that targets are divided
% into training, validation and test sets according to timesteps.
% For a list of data division modes type: help nntype_data_division_mode
net.divideMode = 'value';  % Divide up every value
net.divideParam.trainRatio = 70/100;
net.divideParam.valRatio = 15/100;
net.divideParam.testRatio = 15/100;

% Choose a Training Function
% For a list of all training functions type: help nntrain
% Customize training parameters at: net.trainParam
net.trainFcn = 'trainlm';  % Levenberg-Marquardt

% Choose a Performance Function
% For a list of all performance functions type: help nnperformance
% Customize performance parameters at: net.performParam
net.performFcn = 'mse';  % Mean squared error

% Choose Plot Functions
% For a list of all plot functions type: help nnplot
% Customize plot parameters at: net.plotParam
net.plotFcns = {'plotperform','plottrainstate','plotresponse', ...
  'ploterrcorr', 'plotinerrcorr'};

% Train the Network
[net,tr] = train(net,inputs,targets,inputStates,layerStates);

% Test the Network
outputs = net(inputs,inputStates,layerStates);
errors = gsubtract(targets,outputs);
performance = perform(net,targets,outputs)

% Recalculate Training, Validation and Test Performance
trainTargets = gmultiply(targets,tr.trainMask);
valTargets = gmultiply(targets,tr.valMask);
testTargets = gmultiply(targets,tr.testMask);
trainPerformance = perform(net,trainTargets,outputs)
valPerformance = perform(net,valTargets,outputs)
testPerformance = perform(net,testTargets,outputs)

% View the Network
view(net)

% Plots
% Uncomment these lines to enable various plots.
%figure, plotperform(tr)
%figure, plottrainstate(tr)
%figure, plotregression(targets,outputs)
%figure, plotresponse(targets,outputs)
%figure, ploterrcorr(errors)
%figure, plotinerrcorr(inputs,errors)

% Closed Loop Network
% Use this network to do multi-step prediction.
% The function CLOSELOOP replaces the feedback input with a direct
% connection from the outout layer.
netc = closeloop(net);
netc.name = [net.name ' - Closed Loop'];
view(netc)
[xc,xic,aic,tc] = preparets(netc,inputSeries,{},targetSeries);
yc = netc(xc,xic,aic);
closedLoopPerformance = perform(netc,tc,yc)

% Early Prediction Network
% For some applications it helps to get the prediction a timestep early.
% The original network returns predicted y(t+1) at the same time it is given y(t+1).
% For some applications such as decision making, it would help to have predicted
% y(t+1) once y(t) is available, but before the actual y(t+1) occurs.
% The network can be made to return its output a timestep early by removing one delay
% so that its minimal tap delay is now 0 instead of 1.  The new network returns the
% same outputs as the original network, but outputs are shifted left one timestep.
nets = removedelay(net);
nets.name = [net.name ' - Predict One Step Ahead'];
view(nets)
[xs,xis,ais,ts] = preparets(nets,inputSeries,{},targetSeries);
ys = nets(xs,xis,ais);
earlyPredictPerformance = perform(nets,ts,ys)* | |

Can anyone tell me what I should be doing? How to predict the future values? I don't have any idea how to. Please help me.

I tried using the following code

testinseries = tonndata(testinput,false,false); testtargetSeries = tonndata(testoutput,false,false); netCL = openloop(net); [X,Xi,Ai,T] = preparets(net,testinseries,{},testtargetSeries); y = sim(net,X,Xi,Ai)

to predict the future output, but i am getting values like these.

Columns 1 through 7

    [0.9182]    [0.9161]    [0.9168]    [0.9182]    [0.9181]    [0.9183]    [0.9179]

Columns 8 through 14

    [0.9168]    [0.9170]    [0.9177]    [0.9164]    [0.9176]    [0.9212]    [0.9214]

Columns 15 through 21

    [0.9212]    [0.9195]    [0.9199]    [0.9202]    [0.9190]    [0.9189]    [0.9190]

Columns 22 through 28

    [0.9192]    [0.9217]    [0.9215]    [0.9206]    [0.9221]    [0.9212]    [0.0397]

Columns 29 through 35

    [0.0569]    [0.0569]    [0.0569]    [0.0569]    [0.0569]    [0.0569]    [0.0569]

Columns 36 through 42

    [0.0569]    [0.0569]    [0.0569]    [0.0569]    [0.0570]    [0.0570]    [0.0570]

Columns 43 through 44

    [0.0570]    [0.0570]

From the 28th value, they are the predicted output for the future, where i have given the input and targets as NaN.

I think that since the targets are given as NaN-0 , the values are being predicted like this. I have used closeloop method also, and i am getting similar results. ( i have given the openloop code here)

I really need to complete this project. If i complete this prediction it is over.

Can anyone help me? Please.

Thank you, in advance.

P.S:

Any code snippets would be helpful. If anyone has predicted future values using the tool, please help me.

[ahmed salem]

How can I predict future values of currency prices, in a neural network time series trained with daily data, predicting 3 days into the future from the last day's data.y(t+1)depend on y(t),x(t)and history how predict y(t+2)such that depend on y(t+1),x(t+1)and history but x(t+1)unknown

Answer 1

[Greg Heath]

1. Practice on one or more of the MATLAB timeseries datasets that are the most similar to yours.

 help nndatasets

2. Use as many defaults as possible

3. Omit assignment statements that duplicate default values.

4. Post relevant code and error messages if you have problems.

HTH

Greg

[Charles]

simplenarx_dataset works pretty well, with a bit of tinkering