How to use fitcknn for multiple classes?

1 次查看（过去 30 天）

显示更早的评论

Muhammad Kashif 2018-9-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/421047-how-to-use-fitcknn-for-multiple-classes

评论： Mohd Syamizal Mohd Isa 2020-3-6

在 MATLAB Online 中打开

I am working on facial expression recognition. i made a dataset contain features & classes of 213 images.

Step1: Each row of my dataset represents the features of 1 image. so for 213 images 213 rows
Step2: the last column represents classes like; 1,2,3,4,5,6,7 i used fitcsvm it gives great results but now i want to use knn.

QUESTIONS

How to use fitcknn or any knn classifier to classify
along with cross-validation
and find accuracy precision and recall
help me with this code

clc;
close all;
data = load(fullfile('.', 'Features', 'jaffe_features.txt'));
% features_train = data(1:128,:);
% features_test  = data(128:end,:);
nRows     = size(data,1);
randRows  = randperm(nRows);                    % generate random ordering of row indices
features = data(randRows(1:end),:);
labels1 =  data(:,end);
[labels] = labels1;
Mdl = fitcknn(features,labels,'NumNeighbors',5,...
     'ClassNames',{'1','2','3','4','5','6','7'},'Distance','euclidean', 'Standardize',1);
loss = resubLoss(Mdl);
CVMdl = crossval(Mdl);
classError = kfoldLoss(CVMdl);
label = predict(Mdl,features);
%  plot confusion(features_test,idx)
% oofLabel = kfoldPredict(CVMdl);
%  ConfMat = confusionmat(labels_test,label);
accuracy=confusionmatStats_2(labels_test,label);
% [m,n]=size(label);
% 
% count=0;
% for i=1:m
%     if(strcmp(labels_test(i),label(i)))
%         count=count+1;
%     end
% end
% Regards Regards

2 个评论
显示无隐藏无

fatin suhana mohd khidzir 2019-4-19

hai..i am doing the same knn and svm classifier as yours for facial expression recognition. can you teach me how to classify the 7 facial expression and label it by using knn and svm? can i have your email to learn futher from you? thank you