Nearest Neighbor Algorithm Help!

Hey, so I'm struggling trying to find how to use the nearest neighbor algorithm (yes, NOT the command)in order to organize a group of 2000 points into either class 1 or class 2. I plan on using the norm and sort command to accomplish this, with the input of k determining the size of the neighborhood. Please tell me what I can do to fix this script!! Attached are the data points.

data=load('NN.dat'); [rows,cols]=size('data') class1=data([1:2:500],:),1 class2=data([2:2:500],:),2 x=norm(class1-class2)

if class1>x label=2 end if class1<x label(class1)=1 end if class1>x label(class1)=2 end if class2<x label(class2)=2 end tradata=zeros(1000,3); tradata(:,1:2)=filedata(1:1000,:); tradata([1:2:1000],3)=1; tradata([2:2:1000],3)=2;

Answer 1

[Star Strider]

Here is a KNN classifier code snippet I wrote a few weeks ago. You can adapt it to your situation:

ClassVectors = randi(10, 3, 5);             % Class Vectors (3x5)
DataVectors = randi(10, 15, 5);

for k1 = 1:size(ClassVectors,1)
    for k2 = 1:size(DataVectors,1)
        d(k2,k1) = sqrt(sum((ClassVectors(k1,:)-DataVectors(k2,:)).^2));
    end
end
[~,ClassMember] = min(d, [], 2);

It produces a column vector of class members based on the minimum distance from that vector to the matching class.

[TS]

Thanks, this helps a lot. I greatly appreciate it!

[TS]

Actually, what goes in the brackets?

[Star Strider]

My pleasure!

In the min call, nothing goes in the brackets. That is an empty argument, otherwise min would take the minimum of the first and second arguments rather than just the first, as I want it to here. The third argument instructs min to take the minimum of ‘d’ across columns (dimension #2). Needing to specify the third argument requires an empty argument for the second.