Javascript must be enabled to continue!
Federated Bandit: A Gossiping Approach
View through CrossRef
We study Federated Bandit, a decentralized Multi-Armed Bandit (MAB) problem with a set of N agents, who can only communicate their local data with neighbors described by a connected graph G. Each agent makes a sequence of decisions on selecting an arm from M candidates, yet they only have access to local and potentially biased feedback/evaluation of the true reward for each action taken. Learning only locally will lead agents to sub-optimal actions while converging to a no-regret strategy requires a collection of distributed data. Motivated by the proposal of federated learning, we aim for a solution with which agents will never share their local observations with a central entity, and will be allowed to only share a private copy of his/her own information with their neighbors. We first propose a decentralized bandit algorithm GossipUCB, which is a coupling of variants of both the classical gossiping algorithm and the celebrated Upper Confidence Bound (UCB) bandit algorithm. We show that GossipUCB successfully adapts local bandit learning into a global gossiping process for sharing information among connected agents, and achieves guaranteed regret at the order of O(max(poly(N,M) log T, poly(N,M) logλ2-1 N)) for all N agents, where λ2∈(0,1) is the second largest eigenvalue of the expected gossip matrix, which is a function of G. We then propose FedUCB, a differentially private version of GossipUCB, in which the agents preserve ε-differential privacy of their local data while achieving O(max poly(N,M)/ε log2.5 T, poly(N,M) (logλ2-1 N + log T)) regret.
Association for Computing Machinery (ACM)
Title: Federated Bandit: A Gossiping Approach
Description:
We study Federated Bandit, a decentralized Multi-Armed Bandit (MAB) problem with a set of N agents, who can only communicate their local data with neighbors described by a connected graph G.
Each agent makes a sequence of decisions on selecting an arm from M candidates, yet they only have access to local and potentially biased feedback/evaluation of the true reward for each action taken.
Learning only locally will lead agents to sub-optimal actions while converging to a no-regret strategy requires a collection of distributed data.
Motivated by the proposal of federated learning, we aim for a solution with which agents will never share their local observations with a central entity, and will be allowed to only share a private copy of his/her own information with their neighbors.
We first propose a decentralized bandit algorithm GossipUCB, which is a coupling of variants of both the classical gossiping algorithm and the celebrated Upper Confidence Bound (UCB) bandit algorithm.
We show that GossipUCB successfully adapts local bandit learning into a global gossiping process for sharing information among connected agents, and achieves guaranteed regret at the order of O(max(poly(N,M) log T, poly(N,M) logλ2-1 N)) for all N agents, where λ2∈(0,1) is the second largest eigenvalue of the expected gossip matrix, which is a function of G.
We then propose FedUCB, a differentially private version of GossipUCB, in which the agents preserve ε-differential privacy of their local data while achieving O(max poly(N,M)/ε log2.
5 T, poly(N,M) (logλ2-1 N + log T)) regret.
Related Results
DBA: Dynamic Multi-Armed Bandit Algorithm
DBA: Dynamic Multi-Armed Bandit Algorithm
We introduce Dynamic Bandit Algorithm (DBA), a practical solution to improve the shortcoming of the pervasively employed reinforcement learning algorithm called Multi-Arm Bandit, a...
Federated Data Linkage in Practice
Federated Data Linkage in Practice
In recent years, great strides have been made towards the deployment of federated systems for data research, including exploring federated trusted research environments (TREs). The...
Consequences of Gossiping on Women Empowerment
Consequences of Gossiping on Women Empowerment
Abstract
Gossip is prevalent and is widespread in human society. Gossip has been denigrated as ‘idle talk’, mostly among women based on ‘trifling or groundless ru...
On a Framework for Federated Cluster Analysis
On a Framework for Federated Cluster Analysis
Federated learning is becoming increasingly popular to enable automated learning in distributed networks of autonomous partners without sharing raw data. Many works focus on superv...
One box to search them all
One box to search them all
PurposeThe purpose of this paper is to present how, in May 2008, the Ad Hoc Committee on Federated Search was formed to prepare a preliminary report on federated searching for a sp...
Image-based crop disease detection with federated learning
Image-based crop disease detection with federated learning
Abstract
Crop disease detection and management is critical to improving productivity, reducing costs, and promoting environmentally friendly crop treatment methods. Modern ...
Cloud-Based Federated Learning Implementation Across Medical Centers
Cloud-Based Federated Learning Implementation Across Medical Centers
PURPOSEBuilding well-performing machine learning (ML) models in health care has always been exigent because of the data-sharing concerns, yet ML approaches often require larger tra...
Interactional Styles used by Male Characters when Gossiping at Workplace in The Intern Movie
Interactional Styles used by Male Characters when Gossiping at Workplace in The Intern Movie
The writer conducted this study to investigate the interactional styles used by the male characters, Ben Whittaker, Jason, Davis, and Lewis in The Intern movie. This study was done...

