Dear all;
1: I used SVM to fault detection, now I want to figure out effect of fault in 10 seconds time slot of 60 seconds such that I have a window of data with 10 seconds length like 0-10 seconds,0.005s-10.005 s, 0.01s -10.01s ...60s. and I want to use SVM for each 10 s window of data.
I have some data in excel file which is divided to two parts; upper section is normal data and lower section is fault data and I have 6 features. Entirely I have 24002 data. I wrote a piece of code but I'm not sure if it is correct or not and I want to know how can I correct it? 2:I would like to know how can I divide my data to train and test in for loop for each window?
clc;clear;close all;
T=2001; %length of data in 10s (Window Size)
X=xlsread('Book4');
K=(0.5*length(X))-T+1 % Number of repetitions
window=zeros(2*T,6);
for i=1:K
window=[X(i:i+T-1);X(i+12001:i+12001+T-1,:)];
%% Data Normalaization
m=length(window);
Mean_data=repmat(mean(window),m,1);
Std_data=repmat(std(window),m,1);
data_norm=(window-Mean_data)./(Std_data);
end
I'll appreciate your help.
You can try with this.
clc;clear;close all;
T=2001; %length of data in 10s (Window Size) X=xlsread('Book4');
K=(0.5*length(X))-T+1 % Number of repetitions window=zeros(2*T,6);
for i=1:K window=[X(i:i+T-1);X(i+12001:i+12001+T-1,:)];
%% Data Normalaization
m=length(window); Mean_data=repmat(mean(window),m,1); Std_data=repmat(std(window),m,1); Data_norm=(window-Mean_data)./(Std_data);
end
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory
- Control Systems
- Control Systems Engineering