VowpalWabbit / coba

Contextual bandit benchmarking
https://coba-docs.readthedocs.io/
BSD 3-Clause "New" or "Revised" License
48 stars 19 forks source link

LinUCB Learner #6

Closed anrath closed 3 years ago

anrath commented 3 years ago

Implementing the LinUCB algorithm as specified in Contextual Bandits with Linear Payoff Functions, Chu et. al. Added an interactions parameter to allow for the specification of interactions between context and actions. This term may require future testing and optimization when including higher-order terms.

mrucker commented 3 years ago

This looks great! Nice work.