ballsensitive.m

This code is example for doing sensitivity analysis towards training data

%datai=[0.07865	0.08467	0.31751	-0.30140	1.24090	0.01417	0.00210	3442.15687	4167.07178	1]
%datai=[0.07269	0.07376	0.31125	-0.28664	1.23986	0.02837	0.00097	3028.50500	4167.07178	1]
load allclass4.txt;

load fnet;
datai=allclass4*1.064;
datai1=allclass4*0.95;
[N M]=size(datai);
 nmax=[0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604; 0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604; 0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604;0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604; 0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604; 0.66911	0.66951	6.6529	-0.21819	2.0535	0.12064	0.01222	8260	8604];
 nmin=[0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2; 0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2;0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2;0.06315	0.06433	0.28372	-6.6533	1.2399	0.00999	0.00013	3028.5	4166.2];

% % actual_data = normalise(datai,nmax,nmin);
% %  for i=1:N
% %    nmax(i)=datai(i,N-1);
% %     nmin(i)=datai(i,N-1);
%     for j=1:M
%        if datai(j)>nmax
%           nmax=datai(j)
%         end
%         if datai(j)<nmin
%           nmin=datai(j)
%        end
%     end
% %  end
  % num = data - nmin';
  %dem=nmax - nmin';
   % datae =  num/dem; % no delay
  % for i=1:N
   anu=(datai-nmin);
    anu1=(datai1-nmin);
   ana=(nmax-nmin);
    datas =anu./ana
     datas1 =anu1./ana
    %datad(i,1:M-1)=datai(i,2:M);    % 1 delayed term
  %end  %datad1(i,1:M-2)=datai(i,3:M);    % 2 delayed term
an=sim(fnet,datas') %simulation the network
an1=sim(fnet,datas1')
UL =5;
LL= 3;
figure(i);
output=output*3+1;
an=an*3+1;
an1=an1*3+1;
s=(1:12); %plot the graf data vs month(1986-2001)
plot(s,UL,'r+',s,LL,'r+-',s,an,'b.-',s,an1,'g*-')
xlabel('Observation Data')
ylabel('Classification Output ')
title('Violation output at ball fault condition ')
%legend(' = actual',' = prediction')