Multi-task learning (MTL) represents a paradigm shift in the way machine learning models are developed, whereby multiple related tasks are learned simultaneously. By leveraging shared representations ...

An Osaka Metropolitan University research team conducted a study that proposes a transparent multi-task framework that combines a multi-slot user memory with a dynamic gating network to jointly ...