notesum.ai

Published at November 1

A Similarity-Based Oversampling Method for Multi-label Imbalanced Text Data

cs.LG

cs.AI

Released Date: November 1, 2024

Authors: Ismail Hakki Karaman¹, Gulser Koksal², Levent Eriskin³, Salih Salihoglu¹

Aff.: ¹Department of Industrial Engineering, Middle East Technical University, Ankara, Turkey; ²Department of Industrial Engineering, TED University, Ankara, Turkey; ³Department of Industrial Engineering, Piri Reis University, Istanbul, Turkey

Arxiv: http://arxiv.org/abs/2411.01013v1

Parameter	Values
Balance ratio	0.2, 0.3, 0.4, 0.5
Similarity calculation type	average, safe_interval
Batch size	1, 2, 3, 5, 7
Number of iterations	50, 100, 200, 500
Similarity type	euclidean, cosine, jensen-shannon