Explain the difference between inductive, transductive, and unsupervised Transfer Learning.
Instruction: Provide definitions and examples of each type, highlighting their applications and challenges.
Context: This question requires a deep understanding of the various forms of Transfer Learning, allowing candidates to showcase their comprehensive knowledge and ability to apply these concepts in different scenarios.
Updated
Official answer available
Read the opening below, then unlock the full answer and practical guidance.
Using the terminology of Pan and Yang's survey:
Inductive transfer: the target task differs and target labels are available; source labels may or may not exist. Example: adapt a text representation trained for binary sentiment to a labeled five-star rating task. Replace or adapt the output head and evaluate the target task...