Asymmetric Feature-based Transfer Learning

Homogeneous (where the metrics are the same for both source and target) asymmetric transformation mapping transforms the source feature space to align with that of the target or the target to that of the source. This, in effect, bridges the feature space gap and reduces the problem into a homogeneous transfer problem when further distribution differences need to be corrected.

References

Day, O., & Khoshgoftaar, T.M. (2017). A survey on heterogeneous transfer learning. Journal of Big Data, 4(1), 29. Link.