Using NARX model with Neural Network Prediction

Hello,

Im doing a carbon emission(output) with multiple input using neural network approached (NARX). I trained the model using NARX tollbox. I couldn't find the code I needed to write to predict the next 5 years.

Can you help me?

[Seda]

My code;

% Solve an Autoregression Problem with External Input with a NARX Neural Network
% Script generated by Neural Time Series app
% Created 10-Nov-2023 12:42:04
%
% This script assumes these variables are defined:
%
%   x - input time series.
%   t - feedback time series.
X = tonndata(x,false,false);
T = tonndata(t,false,false);
% Choose a Training Function
% For a list of all training functions type: help nntrain
% 'trainlm' is usually fastest.
% 'trainbr' takes longer but may be better for challenging problems.
% 'trainscg' uses less memory. Suitable in low memory situations.
trainFcn = 'trainlm';  % Levenberg-Marquardt backpropagation.
% Create a Nonlinear Autoregressive Network with External Input
inputDelays = 1:2;
feedbackDelays = 1:2;
hiddenLayerSize = 10;
net = narxnet(inputDelays,feedbackDelays,hiddenLayerSize,'open',trainFcn);
% 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.
[x,xi,ai,t] = preparets(net,X,{},T);
% Setup Division of Data for Training, Validation, Testing
net.divideParam.trainRatio = 80/100;
net.divideParam.valRatio = 10/100;
net.divideParam.testRatio = 10/100;
% Train the Network
[net,tr] = train(net,x,t,xi,ai);
% Test the Network
y = net(x,xi,ai);
e = gsubtract(t,y);
performance = perform(net,t,y)
% View the Network
view(net)
% Plots
% Uncomment these lines to enable various plots.
%figure, plotperform(tr)
%figure, plottrainstate(tr)
%figure, ploterrhist(e)
%figure, plotregression(t,y)
%figure, plotresponse(t,y)
%figure, ploterrcorr(e)
%figure, plotinerrcorr(x,e)
% Closed Loop Network
% Use this network to do multi-step prediction.
% The function CLOSELOOP replaces the feedback input with a direct
% connection from the output layer.
netc = closeloop(net);
netc.name = [net.name ' - Closed Loop'];
view(netc)
[xc,xic,aic,tc] = preparets(netc,X,{},T);
yc = netc(xc,xic,aic);
closedLoopPerformance = perform(net,tc,yc)
% Step-Ahead 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,X,{},T);
ys = nets(xs,xis,ais);
stepAheadPerformance = perform(nets,ts,ys)

Answer 1

[Venu]

Hi @Seda,

I understand you want to predict next 5 years output. For that we need to know format of your input time series dataset. For example you have the data of 10 months, 1 month will be the predicted output time series since your test data is 1/10th. So you need to predict for next 60 months i.e 5 years.

You can find the below example code as reference which is like extention to your code.

% Number of future time steps to predict

numFutureSteps = 60; % 5 years of monthly predictions

% cell array for the predicted values

predictedValues = cell(1, numFutureSteps);

% Initialize the prediction loop state with the last values of the input and layer states

[~, lastInputState, lastLayerState] = preparets(netc, X, {}, T);

lastOutput = yc{end}; % last predicted output from the closed loop network

% predict into the future, one step at a time

for i = 1:numFutureSteps

% use the last predicted output as the current input for the prediction

[nextOutput, nextInputState, nextLayerState] = netc({lastOutput}, lastInputState, lastLayerState);

% store the predicted value

predictedValues{i} = nextOutput;

% Update the last output and states for the next iteration

lastOutput = nextOutput;

lastInputState = nextInputState;

lastLayerState = nextLayerState;

end

% The predictedValues cell array now contains your 5-year predictions

[Seda]

Dear @Venu

Thank you so much...

[Seda]

Dear @Venu

I need help if possible. I want to use the NARX model. I have one input variable and 5 independent variables. I have data for the past 49 years and I want to predict the next 8 years. How can i write code?

Many thanks...

[Seda]

@Venu Also My data is in annual format...

[Seda]

@Venu And i wrote this code;

x=xlsread('Input.xlsx');

t=xlsread('Output.xlsx');

X = tonndata(x,false,false);

T = tonndata(t,false,false);

trainFcn = 'trainlm';

inputDelays = 1:2;

feedbackDelays = 1:2;

hiddenLayerSize = 10;

net = narxnet(inputDelays,feedbackDelays,hiddenLayerSize,'open',trainFcn);

[x,xi,ai,t] = preparets(net,X,{},T);

net.divideParam.trainRatio = 80/100;

net.divideParam.valRatio = 10/100;

net.divideParam.testRatio = 10/100;

[net,tr] = train(net,x,t,xi,ai);

y = net(x,xi,ai);

e = gsubtract(t,y);

performance = perform(net,t,y)

view(net)

netc = closeloop(net);

netc.name = [net.name ' - Closed Loop'];

view(netc)

[xc,xic,aic,tc] = preparets(netc,X,{},T);

yc = netc(xc,xic,aic);

closedLoopPerformance = perform(net,tc,yc)

nets = removedelay(net);

nets.name = [net.name ' - Predict One Step Ahead'];

view(nets)

[xs,xis,ais,ts] = preparets(nets,X,{},T);

ys = nets(xs,xis,ais);

stepAheadPerformance = perform(nets,ts,ys)

numFutureSteps = 8;

predictedValues = cell(1, numFutureSteps);

[~, lastInputState, lastLayerState] = preparets(netc, X, {}, T);

lastOutput = yc{end};

for i = 1:numFutureSteps

[nextOutput, nextInputState, nextLayerState] = netc({lastOutput}, lastInputState, lastLayerState);

predictedValues{i} = nextOutput;

lastOutput = nextOutput;

lastInputState = nextInputState;

lastLayerState = nextLayerState;

[Venu]

Hi @Seda,

Your code seems good to me. Can you elaborate the issue or any error if you are facing with this approach?

Thanks