
Deep Learning for NLP: From the Trenches with Charlene Chambliss - #433
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
00:00
Adapting COVID Classifiers and NLP Innovations
This chapter explores the retraining of a multi-class COVID classifier to address topic drift as the discourse surrounding the pandemic evolved. It details the innovative sampling techniques that enhanced performance metrics and discusses the training methodology for additional models designed to analyze disputed news information. The discussion further emphasizes the importance of effective data labeling and the challenges faced while utilizing different source data types, including news articles and tweets.
Transcript
Play full episode