DocumentCode :
1667587
Title :
High-Order Tensor Decomposition for Large-Scale Data Analysis
Author :
Longzhuang Li ; Boulware, Douglas
Author_Institution :
Sch. of Eng. & Comput. Sci., Air Force Res. Lab. Texas A&M Univ.-Corpus, Christi, TX, USA
fYear :
2015
Firstpage :
665
Lastpage :
668
Abstract :
Higher-order tensor decomposition is a basis for many important data mining tasks and the efficient large-scale tensor decomposition algorithms will have positive impact on clustering, trend detection, and anomaly detection. In the paper, we develop a scalable and distributed version of the Tucker tensor decomposition, MR-T, using the Hadoop MapReduce framework. We avoid large matrix-matrix multiplication and exploit the sparsity of large data sets to minimize intermediate data and flops by sequentially computing the intermediate matrices and generating the intermediate tensor vector-wise.
Keywords :
data analysis; data mining; matrix multiplication; pattern clustering; tensors; Hadoop MapReduce framework; MR-T; Tucker tensor decomposition; anomaly detection; clustering; data mining tasks; high-order tensor decomposition; large data set sparsity; large-scale data analysis; large-scale tensor decomposition algorithms; matrix-matrix multiplication; trend detection; Algorithm design and analysis; Clustering algorithms; Data mining; MATLAB; Matrix decomposition; Signal processing algorithms; Tensile stress; MapReduce-based Tucker decomposition (MR-T); large-scale data analysis; tensor decomposition;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Big Data (BigData Congress), 2015 IEEE International Congress on
Conference_Location :
New York, NY
Print_ISBN :
978-1-4673-7277-0
Type :
conf
DOI :
10.1109/BigDataCongress.2015.104
Filename :
7207288
Link To Document :
بازگشت