A hybrid Transformer-autoencoder plus Trust Score is claimed to detect concept drift earlier and more sensitively than standard autoencoders on synthetic airline data.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Improving Real-Time Concept Drift Detection using a Hybrid Transformer-Autoencoder Framework
A hybrid Transformer-autoencoder plus Trust Score is claimed to detect concept drift earlier and more sensitively than standard autoencoders on synthetic airline data.