Collective Online Learning of Gaussian Processes in Massive Multi-Agent Systems

Hoang, Trong Nghia; Hoang, Quang Minh; Low, Kian Hsiang; How, Jonathan P.

doi:10.1609/aaai.v33i01.33017850

Cited by 22 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a future work, we plan to explore techniques to automatically optimize the distribution P N used by FTS to sample agents by learning the similarity between each agent and the target agent (i.e., the fidelity of each agent). Other than the RFF approximation used in this work, other approximation techniques for GP (such as those based on inducing points [5,6,7,8,18,19,20,21,23,37,38,41,53,57,59,60]) may also be used to derive the parameters to be exchanged between agents, which is worth exploring in future works. Moreover, in our experiments, the hyperparameters of the target agent's GP is learned by maximizing the marginal likelihood; it would be interesting to explore whether the GP hyperparameters can also be shared among the agents, which can potentially facilitate better collaboration.…”

Section: Discussionmentioning

confidence: 99%

Federated Bayesian Optimization via Thompson Sampling

Dai¹,

Low²,

Jaillet³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Bayesian optimization (BO) is a prominent approach to optimizing expensiveto-evaluate black-box functions. The massive computational capability of edge devices such as mobile phones, coupled with privacy concerns, has led to a surging interest in federated learning (FL) which focuses on collaborative training of deep neural networks (DNNs) via first-order optimization techniques. However, some common machine learning tasks such as hyperparameter tuning of DNNs lack access to gradients and thus require zeroth-order/black-box optimization. This hints at the possibility of extending BO to the FL setting (FBO) for agents to collaborate in these black-box optimization tasks. This paper presents federated Thompson sampling (FTS) which overcomes a number of key challenges of FBO and FL in a principled way: We (a) use random Fourier features to approximate the Gaussian process surrogate model used in BO, which naturally produces the parameters to be exchanged between agents, (b) design FTS based on Thompson sampling, which significantly reduces the number of parameters to be exchanged, and (c) provide a theoretical convergence guarantee that is robust against heterogeneous agents, which is a major challenge in FL and FBO. We empirically demonstrate the effectiveness of FTS in terms of communication efficiency, computational efficiency, and practical performance.34th Conference on Neural Information Processing Systems (NeurIPS 2020),

show abstract

Section: Discussionmentioning

confidence: 99%

Federated Bayesian Optimization via Thompson Sampling

Dai¹,

Low²,

Jaillet³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We will also consider our outsourced setting in the active learning context (Cao et al, 2013;Hoang et al, 2014a;Low et al, 2008;2009;Ouyang et al, 2014;Zhang et al, 2016). For applications with a huge budget of function evaluations, we like to couple PO-GP-UCB with the use of distributed/decentralized (Chen et al, 2012;2013a;Hoang et al, 2016;2019b;a;Low et al, 2015;Ouyang & Low, 2018) or online/stochastic (Hoang et al, 2015;Low et al, 2014b;Xu et al, 2014;Teng et al, 2020;Yu et al, 2019a;…”

Section: Discussionmentioning

confidence: 99%

Private Outsourced Bayesian Optimization

Kharkovskii¹,

Dai²,

Low³

2020

Preprint

View full text Add to dashboard Cite

This paper presents the private-outsourced-Gaussian process-upper confidence bound (PO-GP-UCB) algorithm, which is the first algorithm for privacy-preserving Bayesian optimization (BO) in the outsourced setting with a provable performance guarantee. We consider the outsourced setting where the entity holding the dataset and the entity performing BO are represented by different parties, and the dataset cannot be released non-privately. For example, a hospital holds a dataset of sensitive medical records and outsources the BO task on this dataset to an industrial AI company. The key idea of our approach is to make the BO performance of our algorithm similar to that of non-private GP-UCB run using the original dataset, which is achieved by using a random projection-based transformation that preserves both privacy and the pairwise distances between inputs. Our main theoretical contribution is to show that a regret bound similar to that of the standard GP-UCB algorithm can be established for our PO-GP-UCB algorithm. We empirically evaluate the performance of our PO-GP-UCB algorithm with synthetic and real-world datasets.

show abstract

“…Meta learning and model fusion. Meta learning and model fusion, e.g., [15] [23] [51], aim to adapt pre-trained models of related tasks to a new task where synchronized training is not possible. [23] applies knowledge transfer from black-box models to new tasks by learning to capture the correspondence between latent representation that represent the embeddings of different models.…”

Section: Related Workmentioning

confidence: 99%

PubSub-ML: A Model Streaming Alternative to Federated Learning

Gondara

Wang

2023

PoPETs

View full text Add to dashboard Cite

Federated learning is a decentralized learning framework where participating sites are engaged in a tight collaboration, forcing them into symmetric sharing and the agreement in terms of data samples, feature spaces, model types and architectures, privacy settings, and training processes. We propose PubSub-ML, Publish-Subscribe for Machine Learning, as a solution in a loose collaboration setting where each site maintains local autonomy on these decisions. In PubSub-ML, each site is either a publisher or a subscriber or both. The publishers publish differentially private machine learning models and the subscribers subscribe to published models in order to construct customized models for local use, essentially benefiting from other sites' data by distilling knowledge from publishers' models while respecting data privacy. The term “model streaming” comes from the extension of PubSub-ML to decentralized data streams with concept drift. Our extensive empirical evaluation shows that PubSub-ML outperforms federated learning methods by a significant margin.

show abstract

Collective Online Learning of Gaussian Processes in Massive Multi-Agent Systems

Cited by 22 publications

References 12 publications

Federated Bayesian Optimization via Thompson Sampling

Federated Bayesian Optimization via Thompson Sampling

Private Outsourced Bayesian Optimization

PubSub-ML: A Model Streaming Alternative to Federated Learning

Contact Info

Product

Resources

About