Multi-GPU implementation of k-nearest neighbor algorithm

Masek, Jan; Bürget, Radim; Karasek, Jan; Uher, Václav; Dutta, Malay Kishore

doi:10.1109/tsp.2015.7296368

Cited by 6 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…And the performance of k-NN on different datasets under different sceneries has been discussed. Multi-GPU implementation of k-Nearest Neighbor classification algorithm has been proposed in (Masek et al, 2015) keeping big data in mind [size of big data and sources have been discussed in (Rehman et al, 2015). This technique has been shown efficient and effective for large data.…”

Section: Related Workmentioning

confidence: 99%

ck-NN: A Clustered k-Nearest Neighbours Approach for Large-Scale Classification

Ullah

Khan

Emaduddin

2019

ADCAIJ

View full text Add to dashboard Cite

k-Nearest Neighbor (k-NN) is a non-parametric algorithm widely used for the estimation and classification of data points especially when the dataset is distributed in several classes. It is considered to be a lazy machine learning algorithm as most of the computations are done during the testing phase instead of performing this task during the training of data. Hence it is practically inefficient, infeasible and inapplicable while processing huge datasets i.e. Big Data. On the other hand, clustering techniques (unsupervised learning) greatly affect results if you do normalization or standardization techniques, difficult to determine "k" Value. In this paper, some novel techniques are proposed to be used as pre-state mechanism of state-of-the-art k-NN Classification Algorithm. Our proposed mechanism uses unsupervised clustering algorithm on large dataset before applying k-NN algorithm on different clusters that might running on single machine, multiple machines or different nodes of a cluster in distributed environment. Initially dataset, possibly having multi dimensions, is pass through clustering technique (K-Means) at master node or controller to find the number of clusters equal to the number of nodes in distributed systems or number of cores in system, and then each cluster will be assigned to exactly one node or one core and then applies k-NN locally, each core or node in clusters sends their best result and the selector choose best and nearest possible class from all options. We will be using one of the gold standard distributed framework. We believe that our proposed mechanism could be applied on big data. We also believe that the architecture can also be implemented on multi GPUs or FPGA to take flavor of k-NN on large or huge datasets where traditional k-NN is very slow.

show abstract

Section: Related Workmentioning

confidence: 99%

ck-NN: A Clustered k-Nearest Neighbours Approach for Large-Scale Classification

Ullah

Khan

Emaduddin

2019

ADCAIJ

View full text Add to dashboard Cite

show abstract

“…Rocha et al proposed a compact data structure to represent sparse datasets for efficient GPU KNN data representation and distances computation. Masek et al presented a multi‐GPU implementation scalable to four devices which splits train data uniformly into the many devices. The main advantage is that no intercommunication among GPUs is required and therefore using multiple GPUs does not introduce any additional overhead, which is a clear advantage.…”

Section: Data Mining Tasks and Techniquesmentioning

confidence: 99%

A survey on graphic processing unit computing for large‐scale data mining

Cano

2017

WIREs Data Min & Knowl

View full text Add to dashboard Cite

General purpose computation using Graphic Processing Units (GPUs) is a wellestablished research area focusing on high-performance computing solutions for massively parallelizable and time-consuming problems. Classical methodologies in machine learning and data mining cannot handle processing of massive and high-speed volumes of information in the context of the big data era. GPUs have successfully improved the scalability of data mining algorithms to address significantly larger dataset sizes in many application areas. The popularization of distributed computing frameworks for big data mining opens up new opportunities for transformative solutions combining GPUs and distributed frameworks. This survey analyzes current trends in the use of GPU computing for large-scale data mining, discusses GPU architecture advantages for handling volume and velocity of data, identifies limitation factors hampering the scalability of the problems, and discusses open issues and future directions.

show abstract

“…Acceleration by GPU is very promising idea that achieved the significant results in past few years. As mentioned in paper [11], using GPU based training on classifier can speed up calculations of k-NN algorithm up to 750× in comparison with a single CPU version. Using this acceleration can be processed much bigger data.…”

Section: Related Workmentioning

confidence: 99%