Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions

About

Low-rank matrix approximations, such as the truncated singular value decomposition and the rank-revealing QR decomposition, play a central role in data analysis and scientific computing. This work surveys and extends recent research which demonstrates that randomization offers a powerful tool for performing low-rank matrix approximation. These techniques exploit modern computational architectures more fully than classical methods and open the possibility of dealing with truly massive data sets. This paper presents a modular framework for constructing randomized algorithms that compute partial matrix decompositions. These methods use random sampling to identify a subspace that captures most of the action of a matrix. The input matrix is then compressed---either explicitly or implicitly---to this subspace, and the reduced matrix is manipulated deterministically to obtain the desired low-rank factorization. In many cases, this approach beats its classical competitors in terms of accuracy, speed, and robustness. These claims are supported by extensive numerical experiments and a detailed error analysis.

Nathan Halko, Per-Gunnar Martinsson, Joel A. Tropp• 2009

Related benchmarks

TaskDatasetResultRank
Node ClusteringCiteseer
Silhouette Width0.42
36
Node ClusteringPubmed
Silhouette Width0.443
36
Node ClusteringCora
Silhouette Width0.54
36
Information RetrievalSciFact BEIR--
36
Information RetrievalArguana BEIR--
35
Document RetrievalRussian-Wikipedia corpus 10^6 documents
Acc@10.834
15
Oil PredictionCorn
MSE0.3511
12
Protein PredictionCorn
MSE0.463
12
Starch PredictionCorn
MSE0.3508
12
Moisture PredictionCorn
MSE0.1486
12
Showing 10 of 44 rows

Other info

Follow for update