Home

transfer

transfer is a python module implementing tools and methods for transfer learning.

Transfer learning (TL) is an area of machine learning that studies how to leverage knowledge gained from a series of one or more source domains to train a model on a target domain. TL is an interesting topic because there are many situations, in reality, where we have access to a large dataset belonging to a specific domain, and we would like to develop a model to apply to a different situation.

The present wiki contains the documentation of the project.

Contents

• Instance weighting
  • Kernel Mean Matching (KMM).
  • Kullback-Leibler Importance Estimation Procedure (KLIEP).
  • 2-Stage Weighting for Multi-Source Domain Adaptation (2SW-MDA).
• Feature transformation
  • Feature Replication (FDA).