Improved three-way cluster oversampling method—TWKO

Puntos clave

• TWKO algorithm significantly improves classification performance in imbalanced datasets, reducing error rates.
• Empirical evaluations show that the improved method enhances data representation across three clusters for better learning outcomes.
• The enhanced method adapts standard clustering approaches to oversampling, focusing on balancing class distributions effectively.
• This approach may enable more accurate predictions in various applications, but further validation is necessary.