Matlab SVM example - matlab

Question

I am trying to implement SVM for classification. The goal is to output the correct grid of origin of a power signal (.wav file). The grids are titled A-I and there are 93 total signals for the training set and 49 practice signals. I have a 93x10x36 matrix of feature vectors. Does anyone know why I get the errors shown? TrainCorrectGrid and Training_Cepstrum1 both have 93 rows so I don't understand what the problem is. Any help is greatly appreciated.
My code is shown here:
clc; clear; close all;
load('avg_fft_feature (4).mat'); %training feature vectors
load('practice_fft_Mag_all (2).mat'); %practice feauture vectors
load('practice_GridOrigin.mat'); %correct grids of origin for practice data
load PracticeCorrectGrid.mat;
load Training_Cepstrum1;
load Practice_Cepstrum1a;
load fSet1.mat %load in correct practice grids
TrainCorrectGrid=['A';'A';'A';'A';'A';'A';'A';'A';'A';'B';'B';'B';'B';'B';'B';'B';'B';'B';'B';'C';'C';'C';'C';'C';'C';'C';'C';'C';'C';'C';'D';'D';'D';'D';'D';'D';'D';'D';'D';'D';'D';'E';'E';'E';'E';'E';'E';'E';'E';'E';'E';'E';'F';'F';'F';'F';'F';'F';'F';'F';'G';'G';'G';'G';'G';'G';'G';'G';'G';'G';'G';'H';'H';'H';'H';'H';'H';'H';'H';'H';'H';'H';'I';'I';'I';'I';'I';'I';'I';'I';'I';'I';'I'];
%[results,u] = multisvm(avg_fft_feature, TrainCorrectGrid, avg_fft_feature_practice);%avg_fft_feature);
[results,u] = multisvm(Training_Cepstrum1(93,:,1), TrainCorrectGrid, Practice_Cepstrum1a(49,:,1));
disp('Grids of Origin (SVM)');
%Display SVM Results
for i = 1:numel(u)
str = sprintf('%d: %s', i, u(i));
disp(str);
end
%Display Percent Correct
numCorrect = 0;
for i = 1:numel(u)
%if (strcmp(TrainCorrectGrid(i,1), u(i))==1); %compare training to
%training
if (strcmp(PracticeCorrectGrid(i,1), u(i))==1); %compare practice data to training
numCorrect = numCorrect + 1;
end
end
numberOfElements = numel(u);
percentCorrect = numCorrect / numberOfElements * 100;
% percentCorrect = round(percentCorrect, 2);
dispPercent = sprintf('Percent Correct = %0.3f%%', percentCorrect);
disp(dispPercent);
error shown here
The multisvm function is shown here:
function [result, u] = multisvm(TrainingSet,GroupTrain,TestSet)
%Models a given training set with a corresponding group vector and
%classifies a given test set using an SVM classifier according to a
%one vs. all relation.
%
%This code was written by Cody Neuburger cneuburg#fau.edu
%Florida Atlantic University, Florida USA and slightly modified by Renny Varghese
%This code was adapted and cleaned from Anand Mishra's multisvm function
%found at http://www.mathworks.com/matlabcentral/fileexchange/33170-multi-class-support-vector-machine/
u=unique(GroupTrain);
numClasses=length(u);
result = zeros(length(TestSet(:,1)),1);
%build models
for k=1:numClasses
%Vectorized statement that binarizes Group
%where 1 is the current class and 0 is all other classes
G1vAll=(GroupTrain==u(k));
models(k) = svmtrain(TrainingSet,G1vAll);
end
%classify test cases
for j=1:size(TestSet,1)
for k=1:numClasses
if(svmclassify(models(k),TestSet(j,:)))
break;
end
end
result(j) = k;
end
mapValues = 'ABCDEFGHI';
u = mapValues(result);

Answer 1

You state that Training_Cepstrum1 has size [93,10,36]. But when you call multisvm, you are only passing in Training_Cepstrum1(93,:,1) which has size [1,10]. Since TrainCorrectGrid has size [93,1], there is a mismatch in the number of rows.
It looks like you make the same error when passing in Practice_Cepstrum1a.
Try replacing your call to multisvm with
[results,u] = multisvm(Training_Cepstrum1(:,:,1), TrainCorrectGrid, Practice_Cepstrum1a(:,:,1));
This way Training_Cepstrum1(:,:,1) has size [93,10], the same number of rows as TrainCorrectGrid.