A Color-blind 3-Approximation for Chromatic Correlation Clustering and Improved Heuristics

Klodt, Nicolas; Seifert, Lars; Zahn, Arthur; Casel, Katrin; Issac, Davis; Friedrich, Tobias

doi:10.1145/3447548.3467446

Cited by 3 publications

(1 citation statement)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similarly, correlated data affects the error of the selectivity estimators used by query optimizers [19], i.e., highly correlated data leading to higher errors [28]. In machine learning, knowing which features or dimensions are correlated serves in both dimensionality reduction, by pointing at data that can be removed as it does not provide additional information [81], and correlation clustering since the correlation between data is a useful input to clustering algorithms [40]. Finally, correlation computation can be utilized by vector databases to find data vector similarity [57], or in data privacy applications to detect when too much public data leads to information leakage [84].…”

Section: Introductionmentioning

confidence: 99%

AMNES: Accelerating the Computation of Data Correlation Using FPGAs

Chiosa,

Preußer,

Blott

et al. 2023

Proc. VLDB Endow.

View full text Add to dashboard Cite

A widely used approach to characterize input data in both databases and ML is computing the correlation between attributes. The operation is supported by all major database engines and ML platforms. However, it is an expensive operation as the number of attributes involved grows. To address the issue, in this paper we introduce AMNES, a stream analytics system offloading the correlation operator into an FPGA-based network interface card. AMNES processes data at network line rate and the design can be used in combination with smart storage or SmartNICs to implement near data or in-network data processing. AMNES design goes beyond matrix multiplication and offers a customized solution for correlation computation bypassing the CPU. Our experiments show that AMNES can sustain streams arriving at 100 Gbps over an RDMA network, while requiring only ten milliseconds to compute the correlation coefficients among 64 streams, an order of magnitude better than competing CPU or GPU designs.

show abstract