English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Multi-label learning predicts a subset of labels from a given label set foran unseen instance while considering label correlations. A known challenge withmulti-label classification is the long-tailed distribution of labels. Manystudies focus on improving the overall predictions of the model and thus do notprioritise tail-end labels. Improving the tail-end label predictions inmulti-label classifications of medical text enables the potential to understandpatients better and improve care. The knowledge gained by one or moreinfrequent labels can impact the cause of medical decisions and treatmentplans. This research presents variations of concatenated domain-specificlanguage models, including multi-BioMed-Transformers, to achieve two primarygoals. First, to improve F1 scores of infrequent labels across multi-labelproblems, especially with long-tail labels; second, to handle long medical textand multi-sourced electronic health records (EHRs), a challenging task forstandard transformers designed to work on short input sequences. A vitalcontribution of this research is new state-of-the-art (SOTA) results obtainedusing TransformerXL for predicting medical codes. A variety of experiments areperformed on the Medical Information Mart for Intensive Care (MIMIC-III)database. Results show that concatenated BioMed-Transformers outperformstandard transformers in terms of overall micro and macro F1 scores andindividual F1 scores of tail-end labels, while incurring lower training timesthan existing transformer-based solutions for long input sequences.

Improving Predictions of Tail-end Labels using Concatenated BioMed-Transformers for Long Medical Documents