Filters








2 Hits in 1.1 sec

FAST-PCA: A Fast and Exact Algorithm for Distributed Principal Component Analysis [article]

Arpita Gang, Waheed U. Bajwa
2022 arXiv   pre-print
'exact' or 'global' convergence guarantees.  ...  To overcome these aforementioned issues, this paper proposes a distributed PCA algorithm termed FAST-PCA (Fast and exAct diSTributed PCA).  ...  We prove the linear convergence of x (t) g,k to a multiple of q k by proving that x(t) g,k converges to q k at a linear rate.  ... 
arXiv:2108.12373v2 fatcat:ci7ow5em2ve6djvso5xokaxe7m

Distributed Principal Subspace Analysis for Partitioned Big Data: Algorithms, Analysis, and Implementation [article]

Arpita Gang, Bingqing Xiang, Waheed U. Bajwa
2021 arXiv   pre-print
Second, in the case of sample-wise partitioned data, the proposed algorithm and a variant of it are analyzed, and their convergence to the true subspace at linear rates is established.  ...  First, two algorithms are proposed in the paper that can be used for distributed PSA/PCA, with one in the case of data partitioned across samples and the other in the case of data partitioned across (raw  ...  The theoretical guarantees of the S-DOT and SA-DOT algorithms show that our proposed solution has linear convergence rates for the case of a subspace with r > 1, unlike the existing theoretical results  ... 
arXiv:2103.06406v3 fatcat:crmz5xjesvdo7d6kvkp4ae3htq