I'm trying to create a classification system using artificial immune systems but I don't know how to create the loop properly.
1 回表示 (過去 30 日間)
古いコメントを表示
So, I need to create a set of random detectors and then compare them with the real cases using "pdist2". When "pdist2" is under a certain value (0.5), the case is atributed as being part of the respective detector. However, this detector is only valid if it reaches a minimum number of "detections" ((size(percentage,1)/size(X,1))*3).
X=zeros(50,16);
for k=1:size(X,2)
for j=1:size(X,1)
X(j,k)=percentage_min(1,k)+rand(1,1)*(percentage_max(1,k)-percentage_min(1,k));
end
end
D=pdist2(X,percentage);
T=zeros(size(D,1),1);
U=zeros(size(X));
for k=1:size(D,1)
for j=1:size(D,2)
if D(k,j)<0.5
T(k)=T(k)+1;
end
if T(k)>(size(percentage,1)/size(X,1))*3
U(k,:)=X(k,:);
end
end
end
This part of the code is working properly. However, now I need to delete the already atributed cases, create another set of detectors (overwriting X) and use "pdist2" again, until almost all cases are atributed (lets say, per example, while size(percentage,1)>500). When I say to delete a case, I mean to delete a row of "percentage". However, when I do this inside the loop, it never stops. And I also already tried to make the loop backwards but I have the same problem.
while size(percentage,1)>500
for k=1:size(D,1)
for j=1:size(D,2)
if D(k,j)<0.5
T(k)=T(k)+1;
percentage(k,:)=[];
end
if T(k)>(size(percentage,1)/size(X,1))*3
U(k,:)=X(k,:);
end
for m=1:size(X,2)
for n=1:size(X,1)
X(n,m)=percentage_min(1,m)+rand(1,1)*(percentage_max(1,m)-percentage_min(1,m));
end
end
D=pdist2(X,percentage);
T=zeros(size(D,1),1);
end
end
end
This part of the code is not working, but I hope you can understand what I want. Can somebody please help me? And I have 2 other minor problems: when a case is atributed to a detector, it shouldn't be atributed to another one (T should be incremented only once by each case, so total sum should be size(percentage,1)). And when a detector is saved on U, that position shouldn't be overwritten by a new detector.
0 件のコメント
回答 (0 件)
参考
カテゴリ
Help Center および File Exchange で Statistics and Machine Learning Toolbox についてさらに検索
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!