I am classifying data. The data is of high dimensions and seperated in different classes. For each class I have calculated the centroid, which is simply a vector of the mean of each sample. For a better explanation see the simplified example below:
class1:
dimension1 dimension2 dimenson3 mean
sample 1 2 4 6 4
sample 2 8 5 5 3
sample 3 7 9 5 7
centroid= [4 3 7]
now i want to add a new sample (mean 2) to this centroid and calculate the distance of the old centroid to the new centroid
old centroid = [4 3 7]
new centroid = [4 3 7 2]
how can I achieve this (in Matlab)?
Aucun commentaire:
Enregistrer un commentaire