Multi-task learning: A technique that allows a model to learn multiple tasks simultaneously, by training the model with multiple objectives or by sharing the parameters of the model between the tasks.