Algorithm 1. RENN algorithm.

Input: The original data set D.             ${\mathrm{X}}_{\mathrm{i}}$ is the sample in D.

For i = 1,2,…,n

a. Calculate the Euclidean distances between ${\mathrm{X}}_{\mathrm{i}}$ and other samples in D. b. Get the category information of three samples closest to ${\mathrm{X}}_{\mathrm{i}}$. c. If two or more nearest samples’ labels are different from ${\mathrm{X}}_{\mathrm{i}}$, ${\mathrm{X}}_{\mathrm{i}}$ is removed from D. Repeat the above step until ${\mathrm{X}}_{\mathrm{i}}$ cannot be removed. END

Output: The balanced data set ${\mathrm{D}}_{\mathrm{S}}$.