As I am a beginner in MATLAB and started learning on my own. So I am stuck at this piece of code. Can anyone explain to me what is happening in T, Test, and C matrix?
Basically this test.m file is a part of huge code and it is calling multisvm function but I am unable to get the idea about T, test, and C that how it is defined?
T=[ 1 10;2 20;3 30;4 40;5 50;6 66;3 30;4.1 42]; %training matrxi
test=[3 34; 1 14; 2.2 25; 6.2 63]; %test=Testing matrix
C=[1;1;2;2;3;3;2;2]; %C=Group,
results = multisvm(T, C, test);
disp('multi class problem');
disp(results);
Hello Sadia,
It appears that T (train) contains scores/values for a set of cases on some number of variables/features (8 and 2, respectively); C is the correct class to which each case in the T set belongs; and test is a different set of cases and their scores on the same two variables/features. The function attempts to derive rules for matching cases to their class (C) using the training set (T), then applies those rules to the test set and reports the results.
More on the function may be found from the developer's posting: https://www.mathworks.com/matlabcentral/fileexchange/60351-multisvm-trainfeatures-trainlable-testfeatures
Good luck with your work.
What is the multisvm? there is no function called multisvm in Matlab
Your code seems to be taken from
https://www.mathworks.com/matlabcentral/fileexchange/39352-multi-class-svm
Apparently, T, test, and C are matrices. As described, multisvm is a function to run "support vector machine" a supervised machine learning method. I think your problem is the concept and rationale behind it. SVM is a method to partition the data into two classes. There are some assumptions and tips about this method. However, in machine learning, the two steps of training and testing are the bases. We first train the model, then test it, if necessary, calibrate it. Please read some texts in this regard. Best wishes. Babak
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India