0

I am trying to compute the accuracy of the HMAX model. First, I used a simple nearest Neighbor Classifier, secondly the Support Vector Machine (SVM) classifier.

In my laptop, I have the 7.6.0(R2008a) MATLAB version, and I obtained great values of accuracy by using either the nearest neighbor or svm classifer. When I moved to work on my friend's laptop in which the R2013b matlab version is installed instead of 7.6.0(R2008a), I always got the same error by using the svm classifer.

Error in mexSVMTrain (line 1)
function [AlphaY, SVs, Bias, Parameters, nSV, nLabel] = mexSVMTrain(Samples, Labels, Parameters, Weight, Verbose)

Output argument "AlphaY" (and maybe others) not assigned during call to
"C:\Users\xxxx\Documents\MATLAB\aub matlab\osu-svm\mexSVMTrain.m>mexSVMTrain".

Please I need your help in order to resolve this problem. Any help will be very appreciated.

Christina
  • 903
  • 16
  • 39

1 Answers1

1

It's not a solution to your problem. Seems like the existed mexSVMTrain.dll in the old version Matlab needs to be recompiled to work in the new released version, but it is unclear how it is recompiled without any C source code for mexSVMTrain from OSU_SVM.

I would suggest you use libsvm for the SVM part instead.

lennon310
  • 12,503
  • 11
  • 43
  • 61
  • Hello Dear Lennon. Thank you for your answer. I downloaded this library and still I get the same error. – Christina Jan 28 '14 at 15:39
  • Dear Lennon, do you remember when I asked a question in " http://stackoverflow.com/questions/20928389/use-the-osusvm-to-recognize-faces-after-the-c2-layer-in-hmax-model ". I use now my own database, and I don't know why I get different accuracy value at each run. If it is normal, so how can I fix a final result? (for example, in the first run, I got an accuracy=58.0645%, in the second run,I got 61.2903%, in the third run, I got 64.5161%, in the fourth run,I got 61.2903%, in the fifth run I got 70.9677%, in the sixth run,I got 64.5161% ... – Christina Jan 28 '14 at 15:45
  • Do you understand me? can I send you my code via mail to help me? – Christina Jan 28 '14 at 15:46
  • @Christina if you are using libsvm, maybe i can take a look. But I don't use osu_svm. I heard that it has the dll problem as you said in your question else where, and it seems no one could resolve it. Actually the source code of osu_svm seems not accessible, so i'm not sure how they handle your data. If you are using cross validated, probably they shuffle your samples every time in training i guess? – lennon310 Jan 28 '14 at 16:00
  • Hello Dear, Please I asked a question on " http://stackoverflow.com/questions/21429154/using-svm-to-train-my-dataset " and I need a help please :) – Christina Jan 29 '14 at 11:05