Multi-task learning¶
Also known as transfer learning, it is a case when we want to fit many related classification or regression models, and we assume that the input-output mapping is similar across different models.
Hierarchical bayes for multi-taks learning¶
Here we are able to pool information between various regression problems.
Usage¶
Essentially in any case where we want to perform any personalization but we want to draw some strength from the whole distribution.