From: Accurate breast cancer diagnosis using a stable feature ranking algorithm
Notation | Description |
---|---|
S | the value of a stability estimator |
K | the number of nearest neighbors |
(X, y) | a sample of features (X) and its label y |
\(f_{i,k}\) | the index of the \(k^{th}\) feature after the \(i^{th}\) ranking |
m | the number of features in stability analysis |
N | the number of feature ranking experiments |
n | the number of features used for BC diagnosis |
M | the number of BC diagnosis experiments |