A hybrid CNN-transformer-graph network is reported to reach 96% accuracy on LC25000, but the paper lacks architectural detail, code, and a described data split.
What the F-measure doesn't measure: Features, Flaws, Fallacies and Fixes
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
The F-measure or F-score is one of the most commonly used single number measures in Information Retrieval, Natural Language Processing and Machine Learning, but it is based on a mistake, and the flawed assumptions render it unsuitable for use in most contexts! Fortunately, there are better alternatives.
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Multi-Scale Deep Learning for Colon Histopathology: A Hybrid Graph-Transformer Approach
A hybrid CNN-transformer-graph network is reported to reach 96% accuracy on LC25000, but the paper lacks architectural detail, code, and a described data split.