Distributed Regression for Heterogeneous Data Sets

Xing, Yan; Madden, Michael G.; Duggan, Jim; Lyons, Gerard J.

doi:10.1007/978-3-540-45231-7_50

Cited by 9 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Xing et al [10] present such a framework for doing regression in heterogenous datasets. However, these techniques perform poorly as the number of such data partitions increases to millions -as in typical P2P systems.…”

Section: B Related Workmentioning

confidence: 99%

A Scalable Local Algorithm for Distributed Multivariate Regression

BhaduriKanishka

KarguptaHillol

2008

Statistical Analysis

View full text Add to dashboard Cite

This paper offers a local distributed algorithm for multivariate regression in large peer-to-peer environments. The algorithm can be used for distributed inferencing, data compaction, data modeling and classification tasks in many emerging peer-to-peer applications for bioinformatics, astronomy, social networking, sensor networks and web mining. Computing a global regression model from data available at the different peer-nodes using a traditional centralized algorithm for regression can be very costly and impractical because of the large number of data sources, the asynchronous nature of the peer-to-peer networks, and dynamic nature of the data/network. This paper proposes a two-step approach to deal with this problem. First, it offers an efficient local distributed algorithm that monitors the "quality" of the current regression model. If the model is outdated, it uses this algorithm as a feedback mechanism for rebuilding the model. The local nature of the monitoring algorithm guarantees low monitoring cost.Experimental results presented in this paper strongly support the theoretical claims.

show abstract

Section: B Related Workmentioning

confidence: 99%

A Scalable Local Algorithm for Distributed Multivariate Regression

BhaduriKanishka

KarguptaHillol

2008

Statistical Analysis

View full text Add to dashboard Cite

show abstract

“…The basic idea is to learn a model at each site locally (no communication at all) and then, when a new sample comes, predict the output by simply taking an average of the local outputs. Xing et al [30] present such a framework for doing regression in heterogenous datasets. However, these techniques perform poorly as the number of such data partitions increase to millions -as in typical P2P systems.…”

Section: Distributed Multi-variate Regressionmentioning

confidence: 99%

An efficient local Algorithm for Distributed Multivariate Regression in Peer-to-Peer Networks

Bhaduri¹,

Kargupta

2008

Proceedings of the 2008 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

This paper offers a local distributed algorithm for multivariate regression in large peer-to-peer environments. The algorithm is designed for distributed inferencing, data compaction, data modeling and classification tasks in many emerging peer-to-peer applications for bioinformatics, astronomy, social networking, sensor networks and web mining. Computing a global regression model from data available at the different peer-nodes using a traditional centralized algorithm for regression can be very costly and impractical because of the large number of data sources, the asynchronous nature of the peer-to-peer networks, and dynamic nature of the data/network. This paper proposes a two-step approach to deal with this problem. First, it offers an efficient local distributed algorithm that monitors the "quality" of the current regression model. If the model is outdated, it uses this algorithm as a feedback mechanism for rebuilding the model. The local nature of the monitoring algorithm guarantees low monitoring cost. Experimental results presented in this paper strongly support the theoretical claims.

show abstract

“…Therefore, CML can integrate information based on different context of the local sites. The details of the formal and experimental analysis of CML can be referred in our previous publications [2], [3], [5], [4].…”

Section: Context-sensitive Information Integrationmentioning

confidence: 99%

“…However, existing approaches of distributed data mining (DDM) can only deal with the first feature. Therefore, we proposed a contextbased meta-learning (CML) approach to address both of the characteristics [2], [3], [4].…”

Section: Introductionmentioning

confidence: 99%

An Agent-based System for Context-Sensitive Information Integration

Xing

Zhan²

2006 IEEE International Conference on Networking, Sensing and Control

View full text Add to dashboard Cite

Context-based meta-learning (CML) is a new approach for information integration in the application domain of virtual organizations. Aiming at the implementation of CML, this paper designs a multi-agent system to simulate the distributed environment of a virtual organization. Different from others, our system integrates two working modes and adopts two communication mechanisms. The details of system architecture and operation are explained. Experiments are performed to demonstrate that the whole system works well.

show abstract

Distributed Regression for Heterogeneous Data Sets

Cited by 9 publications

References 7 publications

A Scalable Local Algorithm for Distributed Multivariate Regression

A Scalable Local Algorithm for Distributed Multivariate Regression

An efficient local Algorithm for Distributed Multivariate Regression in Peer-to-Peer Networks

An Agent-based System for Context-Sensitive Information Integration

Contact Info

Product

Resources

About