Open access
Aug 2026
WEIGHTED LOSS STRATEGY FOR BERT-BASED TWITTER SENTIMENT ANALYSIS WITHOUT SYNTHETIC OVERSAMPLING
Empirical evidence is provided that, within the present experimental configuration, a properly optimized weighted loss strategy offers a viable and computationally efficient alternative to synthetic oversampling for BERT-based Twitter sentiment classification.
T. Siallagan, R. Winanjaya, Juni Ismail
· JITK (Jurnal Ilmu Pengetahua... · 0 citations