Environment & Distribution Shift
Correction of Distribution Shifts
Covariate Shift Correction
Label Shift Correction
Concept Shift Correction
Concept Shift is much harder to fix in a principled manner.
For instance, in a situation where the problem suddenly changes from
classifying cats v/s dogs to classying white v/s black animals;
it will be unreasonable to assume that we can do much better by collecting new labels and training
from scratch.
Fortunately, in practice such concept-shifts are very rare, but instead it is usually observed that the task keeps on changing gradually. For example:
- Traffic Camera lenses degrade over time due to wear-&-tear and hence the quality of images degrade too.
- News content changes gradually; i.e. new stories appear while the news still remain almost same.
- In computational advertising, new products gets launched and hence popularity of products changes and the CTR-Predictor needs to change as well.
In such above cases, we use the same strategy that we used while trainign the network to make them adapt to change in data i.e.; we use Transfer Learning by finetuning the pre-trained weights with the new aquired data.
A taxonomy of Learning Problems
Armed with the knowledge of dealing with changes in the distribution of data/labels, we now focus on otehr aspects of machine learning formulations: