A bag representation built from averaged Gaussian likelihoods in latent spaces achieves state-of-the-art quantification on the LeQua T1B and T2 multiclass benchmarks.
ISSN: 2374-8486
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Quantification via Gaussian Latent Space Representations
A bag representation built from averaged Gaussian likelihoods in latent spaces achieves state-of-the-art quantification on the LeQua T1B and T2 multiclass benchmarks.