
Dissimilarity of Observations for Sparse Numerical Data
How to measure the dissimilarity of observations for sparse numerical data – a novel approach using weighted symmetric and asymmetric binary distances.
How to measure the dissimilarity of observations for sparse numerical data – a novel approach using weighted symmetric and asymmetric binary distances.
In order to “cluster” observations we need some measure of how alike or different they are. This leads naturally to the concept of a distance between observations. Here we explore how to calculate the distance between observations made up of numerical and ordinal data.