Hi all i am using the fallowing code for the svm classification
svmstruct=svmtrain(PosNegF,group); for l=1:999
abc= svmclassify(svmstruct, DB1(l,:))
Fset(l,:) =[ l,abc]; end here DB is the Feature matrix of the 999 images.
Now when first time i execute this code it work fine, but when next time i train the SVM and then run again this code it give the Subscripted assignment dimension mismatch. error. please help me. Thanks in advance
No products are associated with this question.
3 Comments
Direct link to this comment:
http://www.mathworks.com/matlabcentral/answers/50348#comment_104144
Because, the result or data (svmstruct) will be stored in workspace. To avoid this, you can do with command before this by adding 'clear svmstruct'.
Direct link to this comment:
http://www.mathworks.com/matlabcentral/answers/50348#comment_104145
i try clear svmstruct but still the error is same. please help me..
Direct link to this comment:
http://www.mathworks.com/matlabcentral/answers/50348#comment_104146
At every iteration of loop svmclassify return 1 or 0 only.
then why it giving the dimension mismatch error. even same for the first time.