How to extract the input values for an optimized output value in a GA optimizer

Dear community ,

I am currently working on optimizing a an airofoil

I have an data sheet with 6 elements with 120 samples as an input and 1 element with 120 samples respectively which is the ouput/ target.

Now I have constructed an ANN model with 12 number of hidden neurons in the hidden layer.

The Neural fitting app generates a respective program to provide the matrix only fucntion program which then can be utilized in the GA optimizer tool

Once I save the program and add the progroam as the objective function in the GA optimizer APP with variable as 1

I have an optimized single value for my output data .

But I also need what were the input values to produce the optimized output value.

The below attached program is the Objective function.

function [y1] = myNeuralNetworkFunction1(x1)
%MYNEURALNETWORKFUNCTION neural network simulation function.
%
% Generated by Neural Network Toolbox function genFunction, 16-Nov-2022 11:40:39.
%
% [y1] = myNeuralNetworkFunction(x1) takes these arguments:
%   x = 6xQ matrix, input #1
% and returns:
%   y = 1xQ matrix, output #1
% where Q is the number of samples.

%#ok<*RPMT0>

% ===== NEURAL NETWORK CONSTANTS =====

% Input 1
x1_step1.xoffset = [0;0;0;0;0;0];
x1_step1.gain = [2;2;2;2;2;2];
x1_step1.ymin = -1;

% Layer 1
b1 = [-1.5405239961044094876;-1.5273808353399593862;1.3260615608441381763;-1.5672656908495641304;-0.52463531414582453838;-0.28562547291621864787;-0.92600523410297119753;0.94836491891548213573;0.60413286072586980247;1.3812775835551660553;-0.98746545414147057773;1.8537659554345575774];
IW1_1 = [0.694735700484934382 1.0822554083725968166 0.58162355947880228779 -0.03182512671079117117 -0.35594714793451137647 -0.42115056541279694002;1.7126774333679974927 0.38541323476327671305 0.79165816668606570072 -0.58420613872381599307 0.47446011946542465676 -0.77321888761387003175;-0.47022972931769457805 1.3575902664213252979 -0.022964776251181699684 0.98954132183856580163 -1.3141641243687507412 0.26303920875519415379;0.50570618121209287565 -0.021764214642695456359 -0.17241505384885658092 -1.6282392397915912419 0.35958459771516887438 0.65811302429996787478;0.61453749200090124205 -0.58597896420739203904 0.628444680629517749 0.10933592079163188815 0.64514255935658848529 1.6220061225237449865;-0.93738364435824672594 0.32630035707491461539 -0.99320598793569225826 0.62830932212865575615 -1.1012636140809652918 1.2068758779914174895;-0.36250296717830537974 0.94937094400320753973 0.99760130900678023469 0.70186338551042104505 -1.1900576806232499028 -0.210153503141659731;-0.075593817872469382113 0.24277422296966821857 -1.0955800396372881167 1.2464066543147362953 -0.2124800541687666966 -0.78058215295044663939;0.62073106044722403674 0.12362028693823426395 -0.098384844666925347356 -0.63298401375966018012 0.15725136244761661608 -0.11174225226285876278;0.97877205494941077468 -0.49569157152645215714 -1.7661741788009719389 0.76245089960480827429 -0.875109689249058742 -0.47742261438667982221;-0.50331411621013577573 0.87483246760882549253 -1.2072547488793210491 -0.91121672594314229165 0.77339529988332500476 1.6743937944334255086;1.1929761368367524099 0.99724807116224178927 0.72139890919897775579 0.76024416241570236252 -1.2847855749809071746 -0.19326786300070347702];

% Layer 2
b2 = 0.23670804816455365271;
LW2_1 = [0.16518424966462744163 0.060203722234422205051 0.094778695751439587247 0.48904037181873094564 0.027025074991859011214 -0.042606777963060948888 0.23347377085036169486 0.063412372907630298879 0.41374384009683085051 -0.25859598579246850791 0.034358285747008235345 0.12452529500120101957];

% Output 1
y1_step1.ymin = -1;
y1_step1.gain = 435.824798431031;
y1_step1.xoffset = 0.010163;

% ===== SIMULATION ========

% Dimensions
Q = size(x1,2); % samples

% Input 1
xp1 = mapminmax_apply(x1,x1_step1);

% Layer 1
a1 = tansig_apply(repmat(b1,1,Q) + IW1_1*xp1);

% Layer 2
a2 = repmat(b2,1,Q) + LW2_1*a1;

% Output 1
y1 = mapminmax_reverse(a2,y1_step1);
end

% ===== MODULE FUNCTIONS ========

% Map Minimum and Maximum Input Processing Function
function y = mapminmax_apply(x,settings)
y = bsxfun(@minus,x,settings.xoffset);
y = bsxfun(@times,y,settings.gain);
y = bsxfun(@plus,y,settings.ymin);
end

% Sigmoid Symmetric Transfer Function
function a = tansig_apply(n,~)
a = 2 ./ (1 + exp(-2*n)) - 1;
end

% Map Minimum and Maximum Output Reverse-Processing Function
function x = mapminmax_reverse(y,settings)
x = bsxfun(@minus,y,settings.ymin);
x = bsxfun(@rdivide,x,settings.gain);
x = bsxfun(@plus,x,settings.xoffset);
end

Can you guys please guide me how to overcome the same.

Thanks & Regards.

Answer 1

[Matt J]

You have not shown us the code for the GA app that you are using, but generally speaking, there should be a line in there which calls ga() and which returns the optimum x (and optionally the objective function value fval) that is being iterated over,

[x,fval] = ga(___)

So, maybe you can modify the app code with that in mind.