Spearman's rho¶
Description¶
Spearman's rank correlation coefficient (Spearman's rho) is a non-parametric measure of the strength and direction of a monotonic relationship between two variables. It evaluates how well the relationship between two sets of rankings can be described using a monotonic function. In dimensionality reduction (DR) contexts, it is used to compare the relative order of distances or similarities before and after projection, helping quantify quality loss due to projection.
Formulas¶
Given two variables \(X=(x_1,...,x_n)\) and \(Y=(y_1,...,y_n)\), and their respective ranks \(R(x_i)\) and \(R(y_i)\), Spearman's rho is computed as :
where,
-
\(d_i=R(x_i)-R(y_i)\) is the diference between the ranks of each pair
-
\(n\) is the number of observations
Sources¶
“Applying Deep Learning algorithm to perform lung cells annotation”, A. Collin, 2020