Personalization for web-based services using offline reinforcement learning

Large-scale Web-based services present opportunities for improving UI policies based on observed user interactions. We address challenges of learning such policies through offline reinforcement learning (RL). Deployed in a production system for user authentication in a major social network, it signi...

Full description

Saved in:
Bibliographic Details
Published in:Machine learning 2024-05, Vol.113 (5), p.3049-3071
Main Authors: Apostolopoulos, Pavlos Athanasios, Wang, Zehui, Wang, Hanson, Xu, Tenghyu, Zhou, Chad, Virochsiri, Kittipate, Zhou, Norm, Markov, Igor L.
Format: Article
Language:English
Subjects:
Artificial Intelligence
Computer Science
Control
Costs
Customization
Data collection
Decision making
Distance learning
Machine Learning
Markov analysis
Mechatronics
Natural Language Processing (NLP)
Policies
Recommender systems
Robotics
Simulation and Modeling
Social networks
Special Issue on Reinforcement Learning for Real Life
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Large-scale Web-based services present opportunities for improving UI policies based on observed user interactions. We address challenges of learning such policies through offline reinforcement learning (RL). Deployed in a production system for user authentication in a major social network, it significantly improves long-term objectives. We articulate practical challenges, provide insights on training and evaluation of offline RL, and discuss generalizations toward offline RL’s deployment in industry-scale applications.
ISSN:0885-6125
1573-0565
DOI:10.1007/s10994-024-06525-y