Gamma parameter for SVM (Part 1) | Machine Learning using MATLAB

Explore more about fitcsvm:

Code:
clc
clear all
close all
warning off
load fisheriris
X=meas(:,3:4);
Y=species;
figure
gscatter(X(:,1),X(:,2),Y);
xlabel('Petal Length (cm)');
ylabel('Petal Width (cm)');
classes=unique(Y);
ms=length(classes);
SVMModels=cell(ms,1);
k=0.1;
for j = 1:numel(classes)
indx=strcmp(Y,classes(j)); % Create binary classes for each classifier
SVMModels{j}=fitcsvm(X,indx,'ClassNames',[false true],'Standardize',true,...
'KernelFunction','gaussian','kernelscale',k);
end
e=min(X(:,1)):0.01:max(X(:,1));
f=min(X(:,2)):0.01:max(X(:,2));
[x1 x2]=meshgrid(e,f);
x=[x1(:) x2(:)];
N=size(x,1);
Scores=zeros(N,numel(classes));
for j=1:numel(classes)
[~,score]=predict(SVMModels{j},x);
Scores(:,j)=score(:,2); % Second column contains positive-class scores
end
[~,maxScore]=max(Scores,[],2);
figure
gscatter(x1(:),x2(:),maxScore,'cym');
hold on;
gscatter(X(:,1),X(:,2),Y,'rgb','.',30);
title(k);
xlabel('Petal Length (cm)');
ylabel('Petal Width (cm)');
axis tight
hold off

Learn Machine Learning using MATLAB:

#MachineLearning #MATLAB #DataScience