When 50/50 Isn’t Optimal: Debunking Even Rebalancing | Towards Data Science

A new theory of class imbalance demonstrates that the optimal training imbalance in a binary problem is not 50%

By · · 1 min read
When 50/50 Isn’t Optimal: Debunking Even Rebalancing | Towards Data Science

Source: Towards Data Science

A new theory of class imbalance demonstrates that the optimal training imbalance in a binary problem is not 50%