Fine-tuning pretrained roBERTa model for optimizing relevant biomedical literature search
2021-07-27
·
1 min read
Presented my award-winning machine learning pipeline for the CAMDA 2021 “Literature AI for Drug Induced Liver Injury” (DILI) challenge. Discussed utilizing transformer models (RoBERTa) to automatically classify and extract liver pathology articles from large-scale PubMed datasets.
Authors
Valentyn Bezshapkin
(he/him)
MD, PhD Researcher in Bioinformatics
MD and PhD student in Bioinformatics at ETH Zürich, applying deep learning
and computational methods to decode the human gut microbiome. Building tools
that bridge metagenomics, machine learning, and clinical insights.