Data Consistency for Weakly Supervised Learning

Chidubem Arachie, Bert Huang

In many applications, training machine learning models involves using large amounts of human-annotated data. Obtaining precise labels for the data is expensive. Instead, training w…