Skip to content

Sampling-based methods for correcting for class imbalance in two-category classification problems

License

Notifications You must be signed in to change notification settings

bcbi/ClassImbalance.jl

Repository files navigation

ClassImbalance.jl

DOI

Bors enabled

Description

This is a package that implements some sampling-based methods of correcting for class imbalance in two-category classification problems. Portions of the SMOTE and ROSE algorithm are adaptations of the excellent R packages DMwR and ROSE.

Installation

To install ClassImbalance, open Julia and run the following two lines:

import Pkg
Pkg.add("ClassImbalance")

SMOTE Example

import ClassImbalance;
y = vcat(ones(20), zeros(180)); # 0 = majority, 1 = minority
X = hcat(rand(200, 10), y);
X2, y2 = smote(X, y, k = 5, pct_under = 100, pct_over = 200)