MATLAB Answers

How to get a network function to display input values for best target

1 view (last 30 days)
luke haworth
luke haworth on 28 Mar 2021
Answered: Anshika Chaurasia on 15 Apr 2021
hi all, my project involves collecting data for horsepower results with 5 input vairiables (intake length, intake valve opening time/closing time and exhaust valve closing time)
i have created a feedforward net to lean the relation between the 5 input vairiables and the horsepower results and im happy that the network works well.
what i now need to do is ask the network what input vairiable values i need to get the highest horsepower result
the upper limits of my input data for length, intake open time, intake close time, exhaust open time and exhaust close time is 450,25,70,50 and 30 respectivly
the lower limits of my input data for length, intake open time, intake close time, exhaust open time and exhaust close time is 250,-5,40,20 and 0 respectivly
here is my code for the network
%network for 6000 rpm
filename4 = 'data rearranged.xlsx';
%x_6000 = length
x_6000 = xlsread(filename4,'ADU3:AGV3');
%x2_6000 = IVO timing
x2_6000 = xlsread(filename4,'ADU4:AGV4');
%x3_6000 = IVC timing
x3_6000 = xlsread(filename4,'ADU5:AGV5');
%x4_6000 = EVO timing
x4_6000 = xlsread(filename4,'ADU6:AGV6');
%x5_6000 = EVC timing
x5_6000 = xlsread(filename4,'ADU7:AGV7');
x6000 = [x_6000;x2_6000;x3_6000;x4_6000;x5_6000];
%y1_6000 = hp_6000
y1_6000 = xlsread(filename4,'ADU8:AGV8');
net_hp_6000 = feedforwardnet(10);
net_hp_6000 = configure(net_hp_6000,x6000,y1_6000);
[net_hp_6000,tr] = train(net_hp_6000,x6000,y1_6000);
genFunction(net_hp_6000,'network_horsepower_6000rpm');

Answers (1)

Anshika Chaurasia
Anshika Chaurasia on 15 Apr 2021
Hi,
Feature selection is a dimensionality reduction technique that selects a subset of features (predictor variables) that provide the best predictive power in modeling a set of data.
Refer to the following links for more information of feature selection and available methods in MATLAB:
Hope it helps!

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!