Distributed Quantile Regression Over Sensor Networks

Wang, Heyu; Li, Chunguang

doi:10.1109/tsipn.2017.2699923

Cited by 32 publications

(27 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We focus on parametric regression analytics, e.g., [2], [16], [14] in a (d+1)-dimensional data space (x, y) ∈ R d+1 , where we seek to learn the dependency between input x and output y estimated by the unknown global data function y = f (x) :…”

Section: Rationale and Problem Fundamentalsmentioning

confidence: 99%

See 1 more Smart Citation

Edge-Centric Efficient Regression Analytics

Harth

Anagnostopoulos

2018

2018 IEEE International Conference on Edge Computing (EDGE)

View full text Add to dashboard Cite

We introduce an edge-centric parametric predictive analytics methodology, which contributes to real-time regression model caching and selective forwarding in the network edge where communication overhead is significantly reduced as only model's parameters and sufficient statistics are disseminated instead of raw data obtaining high analytics quality. Moreover, sophisticated model selection algorithms are introduced to combine diverse local models for predictive modeling without transferring and processing data at edge gateways. We provide mathematical modeling, performance and comparative assessment over real data showing its benefits in edge computing environments.

show abstract

Section: Rationale and Problem Fundamentalsmentioning

confidence: 99%

“…Regression includes parametric and non-parametric approaches [16], [20], [17]. Non-parametric approaches use stored data X for predictions, which in our context, are not computationally efficient in terms of data storage, calculations, and on-line updates/adaptations to incoming data [17].…”

Section: A Problem Formulationmentioning

confidence: 99%

Edge-Centric Efficient Regression Analytics

Harth

Anagnostopoulos

2018

2018 IEEE International Conference on Edge Computing (EDGE)

View full text Add to dashboard Cite

show abstract

“…Thus agents must exchange information with their neighbors to find an optimal solution. The motivating examples include formation control [1], [2], large scale machine learning [3], [4], and distributed quantile regression over sensor networks [5]. An overview of this topic can be found in [6].…”

Section: Introductionmentioning

confidence: 99%

“…We approach that problem using our framework and theory, and confirm that the distributed quantile regression can be well solved using only sign of relative state. Compared with [5], the feedback information from each neighbor is now reduced to essentially only one bit at every node.…”

Section: Introductionmentioning

confidence: 99%

Distributed Discrete-Time Optimization in Multiagent Networks Using Only Sign of Relative State

Zhang

You

Başar

2019

IEEE Trans. Automat. Contr.

View full text Add to dashboard Cite

This paper proposes distributed discrete-time algorithms to cooperatively solve an additive cost optimization problem in multi-agent networks. The striking feature lies in the use of only the sign of relative state information between neighbors, which substantially differentiates our algorithms from others in the existing literature. We first interpret the proposed algorithms in terms of the penalty method in optimization theory and then perform non-asymptotic analysis to study convergence for static network graphs. Compared with the celebrated distributed subgradient algorithms, which however use the exact relative state information, the convergence speed is essentially not affected by the loss of information. We also study how introducing noise into the relative state information and randomly activated graphs affect the performance of our algorithms. Finally, we validate the theoretical results on a class of distributed quantile regression problems.Index Terms-Distributed optimization, multi-agent networks, sign of relative state, penalty method, subgradient iterations. arXiv:1709.08360v3 [cs.SY]

show abstract

“…For example, the memory of a personal computer only has a storage size in GBs while the dataset on the hard disk could have a much larger size. In addition, in a sensor network, each sensor is designed to collect and store a limited amount of data, and computations are performed via communications and aggregations among sensors (see, e.g., Wang and Li (2018)). Other examples include high-speed data streams that are transient and arrive at the processor at a high speed.…”

mentioning

confidence: 99%

Quantile regression under memory constraint

Chen¹,

Liu²,

Zhang³

2019

Ann. Statist.

123

View full text Add to dashboard Cite

This paper studies the inference problem in quantile regression (QR) for a large sample size n but under a limited memory constraint, where the memory can only store a small batch of data of size m. A natural method is the naïve divide-and-conquer approach, which splits data into batches of size m, computes the local QR estimator for each batch, and then aggregates the estimators via averaging. However, this method only works when n = o(m 2 ) and is computationally expensive. This paper proposes a computationally efficient method, which only requires an initial QR estimator on a small batch of data and then successively refines the estimator via multiple rounds of aggregations. Theoretically, as long as n grows polynomially in m, we establish the asymptotic normality for the obtained estimator and show that our estimator with only a few rounds of aggregations achieves the same efficiency as the QR estimator computed on all the data. Moreover, our result allows the case that the dimensionality p goes to infinity. The proposed method can also be applied to address the QR problem under distributed computing environment (e.g., in a large-scale sensor network) or for real-time streaming data.

show abstract

Distributed Quantile Regression Over Sensor Networks

Cited by 32 publications

References 31 publications

Edge-Centric Efficient Regression Analytics

Edge-Centric Efficient Regression Analytics

Distributed Discrete-Time Optimization in Multiagent Networks Using Only Sign of Relative State

Quantile regression under memory constraint

Contact Info

Product

Resources

About