K-Means Clustering (KMC) Racial Fallacy:Race:
is a group of similar creatures, categorised by common ancestry.It is not a culture, identity, etc that can be adopted, rejected, identified as or as not; it is biological and unchangeable.
K-Means Clustering: is a method of vector quantization, that aims to partition
n observations into
k clusters in which each observation belongs to the cluster with
the nearest mean.
It is
not an accurate tool in generating true racial groups, and will mis-group large amounts of data, as it divides the dataset evenly into
k groups and attributes everything in a range to the nearest
mean, which makes real races blur across clusters.
KMC is great/ useful for a quick general division, however is especially inaccurate at cluster boundaries: identifying a West Eurasian group which diverged from it's mean and clustered with the unrelated East Eurasian mean, as East Eurasian, which falsely labels half of Europe as partially Asian.
Race is common descent, not dataset clusters.
