Federated Contextual Cascading Bandits with Asynchronous Communication and Heterogeneous Users (2402.16312v1)
Abstract: We study the problem of federated contextual combinatorial cascading bandits, where $|\mathcal{U}|$ agents collaborate under the coordination of a central server to provide tailored recommendations to the $|\mathcal{U}|$ corresponding users. Existing works consider either a synchronous framework, necessitating full agent participation and global synchronization, or assume user homogeneity with identical behaviors. We overcome these limitations by considering (1) federated agents operating in an asynchronous communication paradigm, where no mandatory synchronization is required and all agents communicate independently with the server, (2) heterogeneous user behaviors, where users can be stratified into $J \le |\mathcal{U}|$ latent user clusters, each exhibiting distinct preferences. For this setting, we propose a UCB-type algorithm with delicate communication protocols. Through theoretical analysis, we give sub-linear regret bounds on par with those achieved in the synchronous framework, while incurring only logarithmic communication costs. Empirical evaluation on synthetic and real-world datasets validates our algorithm's superior performance in terms of regrets and communication costs.
- Improved algorithms for linear stochastic bandits. Advances in neural information processing systems, 24.
- Improved Algorithms for Linear Stochastic Bandits. In Proceedings of the 24th International Conference on Neural Information Processing Systems, NIPS’11, 2312–2320. Red Hook, NY, USA: Curran Associates Inc. ISBN 9781618395993.
- The k-means algorithm: A comprehensive survey and performance evaluation. Electronics, 9(8): 1295.
- Cascading Contextual Assortment Bandits. In Thirty-seventh Conference on Neural Information Processing Systems.
- Click models for web search. Synthesis lectures on information concepts, retrieval, and services, 7(3): 1–115.
- An experimental comparison of click position-bias models. In Proceedings of the 2008 international conference on web search and data mining, 87–94.
- Differentially-private federated linear bandits. Advances in Neural Information Processing Systems, 33: 6003–6014.
- Online clustering of bandits. In International Conference on Machine Learning, 757–765. PMLR.
- A Simple and Provably Efficient Algorithm for Asynchronous Federated Contextual Linear Bandits. Advances in neural information processing systems.
- Distributed clustering of linear bandits in peer to peer networks. In International conference on machine learning, 1301–1309. PMLR.
- Cascading bandits: Learning to rank in the cascade model. In International Conference on Machine Learning, 767–776. PMLR.
- Combinatorial cascading bandits. In Proceedings of the 28th International Conference on Neural Information Processing Systems-Volume 1, 1450–1458.
- Asynchronous Upper Confidence Bound Algorithms for Federated Linear Bandits. arXiv preprint arXiv:2110.01463.
- Improved Algorithm on Online Clustering of Bandits. In Proceedings of the 28th International Joint Conference on Artificial Intelligence, IJCAI’19, 2923–2929. AAAI Press. ISBN 9780999241141.
- Contextual combinatorial cascading bandits. In International conference on machine learning, 1245–1253. PMLR.
- Online clustering of contextual cascading bandits. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 32.
- Training Recommenders Over Large Item Corpus With Importance Sampling. IEEE Transactions on Knowledge and Data Engineering.
- Personalized Ranking with Importance Sampling. In Proceedings of The Web Conference 2020, 1093–1103.
- Federated Online Clustering of Bandits. In The 38th Conference on Uncertainty in Artificial Intelligence.
- Batch-size independent regret bounds for combinatorial semi-bandits with probabilistically triggered arms or independent arms. Advances in Neural Information Processing Systems, 35: 14904–14916.
- Contextual combinatorial bandits with probabilistically triggered arms. In International Conference on Machine Learning, 22559–22593. PMLR.
- Contextual Combinatorial Bandits with Probabilistically Triggered Arms. ArXiv, abs/2303.17110.
- Variance-adaptive algorithm for probabilistic maximum coverage bandits with general feedback. In IEEE INFOCOM 2023-IEEE Conference on Computer Communications, 1–10. IEEE.
- Minimax Regret for Cascading Bandits. arXiv preprint arXiv:2203.12577.
- Distributed Bandit Learning: Near-Optimal Regret with Efficient Communication. arXiv e-prints.
- Online clustering of bandits with misspecified user models. arXiv preprint arXiv:2310.02717.
- Influence-Driven Data Poisoning for Robust Recommender Systems. IEEE Transactions on Pattern Analysis and Machine Intelligence.
- Decentralized Randomly Distributed Multi-agent Multi-armed Bandit with Heterogeneous Rewards. In Advances on Neural Information Processing Systems.
- Federated bandit: A gossiping approach. In Abstract Proceedings of the 2021 ACM SIGMETRICS/International Conference on Measurement and Modeling of Computer Systems, 3–4.
- Cascading bandits for large-scale recommendation problems. arXiv preprint arXiv:1603.05359.