Parallel K-means Algorithm Research Articles

A MapReduce-based approach to social network big data mining

The rapid development of social networks has facilitated the convenience of users to receive information. As a network communication platform for people’s daily use, microblog has countless information data. In view of the low efficiency and poor clustering effect of K-means algorithm, a parallel K-means clustering algorithm based on MapReduce model is studied; In order to alleviate the difficulty in calculating the similarity of microblog topic text, the space vector model and semantic similarity are used to calculate the similarity between texts to improve the quality of microblog text classification. The data expansion rate of corresponding nodes under different data sets shows that the average expansion rate of the parallel K-means algorithm reaches 0.89, and the running rate is the highest. The results show that the parallel K-means algorithm has good clustering stability and the highest clustering quality, reaching 1.24; The clustering time of the algorithm is the shortest, the average clustering time is 1.27 minutes, and the clustering effect and efficiency of the algorithm are the best. In the quality analysis of Weibo topic recommendation, the accuracy of P-K-means recommendation is 95.64%, user satisfaction is 98.64%, and the recommendation effect is also the best. It shows that the research on the parallel K-means clustering algorithm based on MapReduce has the best performance in microblogging topic mining and recommendation, which can efficiently recommend topics of interest to users and enhance users’ microblogging experience.

Analysis of Internet Marketing Forecast Model Based on Parallel K-Means Algorithm

Based on the parallel K-means algorithm, this article conducts in-depth research on the related issues of marketing node detection under the Internet, including designing a new Internet marketing node detector and a location summary network based on FCN (Full Convolutional Network) to input the preprocessing of the node and verify its performance under the data sets. At the same time, to solve the problem of insufficient data sets of Internet marketing nodes, the Internet data sets are artificially generated and used for detector training. First, the multiclass K-means algorithm is changed to two categories suitable for Internet marketing node detection: marketing nodes and background categories. Secondly, the weights in the K-means algorithm are mostly only applicable to target detection tasks. Therefore, when processing Internet marketing node detection tasks, the K-means algorithm is used to regress the training set and calculate 5 weights. During the simulation experiment, the weight calculation formula is used to calculate the weight of the feature term. The basic idea is that if a feature word appears more often in this document but less frequently in other nodes, the word will be assigned higher. At the same time, this article focuses on k. Some shortcomings of the mean clustering algorithm have been specifically improved. By standardizing the data participating in the clustering, the data participating in the clustering is transformed from an irregular distribution to a cluster-like distribution, thereby facilitating the clustering process. The density is introduced to determine the initial center of the cluster, and the purity metric is introduced to determine the appropriate density radius of the cluster center, to achieve the most effective reduction of the support vector machine training samples.

A Parallel Clustering Analysis Based on Hadoop Multi-Node and Apache Mahout

The conventional procedures of clustering algorithms are incapable of overcoming the difficulty of managing and analyzing the rapid growth of generated data from different sources. Using the concept of parallel clustering is one of the robust solutions to this problem. Apache Hadoop architecture is one of the assortment ecosystems that provide the capability to store and process the data in a distributed and parallel fashion. In this paper, a parallel model is designed to process the k-means clustering algorithm in the Apache Hadoop ecosystem by connecting three nodes, one is for server (name) nodes and the other two are for clients (data) nodes. The aim is to speed up the time of managing the massive scale of healthcare insurance dataset with the size of 11 GB and also using machine learning algorithms, which are provided by the Mahout Framework. The experimental results depict that the proposed model can efficiently process large datasets. The parallel k-means algorithm outperforms the sequential k-means algorithm based on the execution time of the algorithm, where the required time to execute a data size of 11 GB is around 1.847 hours using the parallel k-means algorithm, while it equals 68.567 hours using the sequential k-means algorithm. As a result, we deduce that when the nodes number in the parallel system increases, the computation time of the proposed algorithm decreases.

Study on academic Early Warning System for Information Engineering Vocational students

Due to the expansion of higher vocational college enrollment, the different quality of students and the high quality requirements of enterprise talents, this paper proposes an academic early warning system for higher vocational college students, which can be used for the early warning of students’ performance and improve the teaching quality. Firstly, the clustering algorithm is analyzed, and the parallel K-means algorithm is adopted in the general algorithm. During the startup and running of the program, the processing time of the algorithm is short and the total consumption of the system is small. Compared with the existing data mining technology, parallel K-means algorithm has good practicability and flexibility.

An Efficient Clustering Based CBIR System using Hadoop to Analyze Massive MRI Images Dataset for Early Disease Diagnosis

In past decade, the use of medical imaging for disease diagnosis is increased rapidly. The medical images provide useful information about the anatomy of patients. The medical images are used not only to assist the doctors for diagnosis purpose, but also used in Research & Development for deeper insights and better understanding into cause and cure of numerous diseases. To retrieve medical images from large scale repositories in real time, there is urgent need of an efficient medical image retrieval system. For this purpose, an efficient clustering based content based image retrieval (CBIR) system using Hadoop is proposed to analyze massive magnetic resonance imaging (MRI) images dataset for early disease diagnosis. The proposed CBIR system uses Hadoop platform, local mesh peak valley edge patterns (LMePVEP) for feature extraction, MapReduce based parallel k-means algorithm for clustering and euclidean distance to measure similarity. The method proposed is tested and compared with state-of-the-art CBIR methods on massive MRI images dataset. The experimental results obtained show that the method proposed in our work outperforms traditional CBIR methods in terms of average retrieval time and mean average precision for massive MRI images dataset.

Enabling Highly Efficient k-Means Computations on the SW26010 Many-Core Processor of Sunway TaihuLight

With the advent of the big data era, the amounts of sampling data and the dimensions of data features are rapidly growing. It is highly desired to enable fast and efficient clustering of unlabeled samples based on feature similarities. As a fundamental primitive for data clustering, the k-means operation is receiving increasingly more attentions today. To achieve high performance k-means computations on modern multi-core/many-core systems, we propose a matrix-based fused framework that can achieve high performance by conducting computations on a distance matrix and at the same time can improve the memory reuse through the fusion of the distance-matrix computation and the nearest centroids reduction. We implement and optimize the parallel k-means algorithm on the SW26010 many-core processor, which is the major horsepower of Sunway TaihuLight. In particular, we design a task mapping strategy for load-balanced task distribution, a data sharing scheme to reduce the memory footprint and a register blocking strategy to increase the data locality. Optimization techniques such as instruction reordering and double buffering are further applied to improve the sustained performance. Discussions on block-size tuning and performance modeling are also presented. We show by experiments on both randomly generated and real-world datasets that our parallel implementation of k-means on SW26010 can sustain a double-precision performance of over 348.1 Gflops, which is 46.9% of the peak performance and 84% of the theoretical performance upper bound on a single core group, and can achieve a nearly ideal scalability to the whole SW26010 processor of four core groups. Performance comparisons with the previous state-of-the-art on both CPU and GPU are also provided to show the superiority of our optimized k-means kernel.

Optimal Tabu K-Means Clustering Using Massively Parallel Architecture

Parallel discovery of inherent clusters using massively threaded architectures is the solution for handling computational challenges raised by fat datasets in cluster analysis. The Graphics Processing Unit and Compute Unified Device Architecture form a convincing platform to parallelize clustering algorithms. The parallel K-means algorithm aims at increasing the speedup, but often faces the hitch of falling into local minima. The heuristic search procedure to discover the global optima in the solution space is known as Tabu Search. The K-means clustering solution is fine-tuned by applying parallel implementation of Tabu Search K-means clustering in order to increase efficacy. The aim is to combine optimization characteristic of Tabu Search for calculation of centroids with clustering attitude of K-means and to enhance the solution using processing power of GPU. The parallelization strategy used exhibits increase in speedup. The parallel Tabu-KM algorithm is tested on standard datasets, and performance is compared with sequential K-means, parallel K-means and sequential Tabu K-means algorithms. The experimental results confirm yet another parallelization technique to unravel data clustering problems.

An analysis of MapReduce efficiency in document clustering using parallel K-means algorithm

One of the significant data mining techniques is clustering. Due to expansion and digitalization of each field, large datasets are being generated rapidly. Such large dataset clustering is a challenge for traditional sequential clustering algorithms due to huge processing time. Distributed parallel architectures and algorithms are thus helpful to achieve performance and scalability requirement of clustering large datasets. In this study, we design and experiment a parallel k-means algorithm using MapReduce programming model and compared the result with sequential k-means for clustering varying size of document dataset. The result demonstrates that proposed k-means obtains higher performance and outperformed sequential k-means while clustering documents.

A GPU-accelerated parallel K-means algorithm

Clustering approaches are widely used methodologies to analyse large data sets. The K-means algorithm is well-known as a procedure too computational-intensive for the large data analytic problem. In this work, we focus on a parallel technique to reduce the execution time when the K-means is used to cluster large dataset. We exploit computational powerful of its design when the Graphic Processor Units (GPUs), a massively parallel architecture, is adopted. We optimize the proposed implementation to handle (i) the space limitation issue of GPUs; (ii) the host-device data transfer time. Experimental results, on real and synthetic data, show how our parallelization approach give good results in terms of execution time and speed-up.

An improved parallel K-means algorithm based on MapReduce

An improved parallel K-means algorithm based on MapReduce

The Parallel Implementation and Application of an Improved K-means Algorithm

Following with the growth of massive data, clustering research, one of the core problems of big dataisfaced with more and more problems such as high computing complexity and lack of resource. It has proposed an improved parallel K-means algorithm based on Hadoop. To overcomethe problem that the traditional K-means algorithm often has local optimal solution due to the randomness choice of initial center, we introduce Canopy algorithm to initialize clustering center andapply K-means algorithm on canopy. Meanwhile, clusters are merged among canopies. The result is stable and iteration number is less. In addition, the parallel implementation methods and strategies of the improved algorithm are presented, combining with the distributed computing model of MapReduce. And a new method of text clustering is introduced by improving the similarity of measurement. The experiment results indicate the validity and scalability of our method.

An improved parallel K-means algorithm based on MapReduce

The K-means algorithm is one of the most popular clustering algorithms. However, it is sensitive to initialised partitions and circular dataset. To address this problem, this paper introduces a CK-means clustering algorithm based on the K-means algorithm and the Canopy algorithm, which uses the MapReduce programming model of Hadoop platform. The experimental results prove that the CK-means algorithm has strong advantages for processing large datasets. The theoretical analysis shows that the CK-means algorithm and the traditional algorithm are of the same order of magnitude. The experimental results on artificial data show that the improved algorithm is better than the traditional algorithm in terms of acceleration ratio, accuracy and expansion rate. An experiment on real data is performed to obtain appropriate parameters.

Dynamic parallel K-Means Algorithm Based On Dunn’s Index Method

Clustering can be define as a method of unsupervised classification, in which the data points are grouped into cluster based on their similarity. K-means is an efficient and widely used algorithm for the purpose of partitioned clustering. K-means algorithm is effective in producing clusters for many real time practical applications. In this Work a modified parallel K-Means algorithm is proposed that overcome the problem of fixed number of input clusters and their serial execution. Originally K-means algorithm accepts the fix number of clusters as input. But in real world scenario it is very difficultto fix the number of clusters in order to get the optimal outcomes. K-means is popular and widely used clustering technique. Many research has been already done in same area for the improvement of K-means clustering, but further investigation is always required to reveal the answers of the important questions such as 'is it possible to find optimal number of clusters dynamically while ignoring the empty clusters' or 'does the parallel execution of any clustering algorithm really improves it performance in terms of speedup'. This research presents an improved K-Means algorithm which is capable to calculate the number of clusters dynamically using Dunn's index approach and further executes the algorithm in parallel using the capabilities of Microsoft's Task Parallel Libraries. The original K-Means and Improved parallel modified K-Means algorithm performed for the two dimensional raw data consisting different numbers of records. From the results it is clear that the Improved K-Means is better in all the scenarios either increase the numbers of clusters or change the number of records in raw data. For the same number of input clusters and different data sets in original K-Means and Improved K-Means, the performance ofModified parallel K-Means is 18 to 46 percent better than the original K-Means in terms of Execution time and Speedup. IndexTerms—K-Means, Parallel K-Means,Task Parallel Libraries, Dunn's index.

A New Approach for Large-Scale Scene Image Retrieval Based on Improved Parallelk-Means Algorithm in MapReduce Environment

The rapid growth of digital images has caused the traditional image retrieval technology to be faced with new challenge. In this paper we introduce a new approach for large-scale scene image retrieval to solve the problems of massive image processing using traditional image retrieval methods. First, we improved traditionalk-Means clustering algorithm, which optimized the selection of the initial cluster centers and iteration procedure. Second, we presented a parallel design and realization method for improvedk-Means algorithm applied it to feature clustering of scene images. Finally, a storage and retrieval scheme for large-scale scene images was put forward using the large storage capacity and powerful parallel computing ability of the Hadoop distributed platform. The experimental results demonstrated that the proposed method achieved good performance. Compared with the traditional algorithms with single node architecture and parallelk-Means algorithm, the proposed method has obvious advantages for use in large-scale scene image data retrieval in terms of retrieval accuracy, retrieval time overhead, and computational performance (speedup and efficiency, sizeup, and scaleup), which is a significant improvement from applying parallel processing to intelligent algorithms with large-scale datasets.

A comparative study of the parallel wavelet-based clustering algorithm on three-dimensional dataset

Cluster analysis--as a technique for grouping a set of objects into similar clusters--is an integral part of data analysis and has received wide interest among data mining specialists. The parallel wavelet-based clustering algorithm using discrete wavelet transforms has been shown to extract the approximation component of the input data on which objects of the clusters are detected based on the object connectivity property. However, this algorithm suffers from inefficient I/O operations and performance degradation due to redundant data processing. We address these issues to improve the parallel algorithm's efficiency and extend the algorithm further by investigating two merging techniques (both merge-table and priority-queue based approaches), and apply them on three-dimensional data. In this study, we compare two parallel WaveCluster algorithms and a parallel K-means algorithm to evaluate the implemented algorithms' effectiveness.

An Investigation of Parallel Road Map Inference from Big GPS Traces Data

Abstract With the increased use of GPS sensors in several everyday devices, persons trip data are be- coming very abundant. Many opportunities for exploration of the wealth GPS data and in this paper, we inferred, the geometry of road maps in Tunisia and the connectivity between them. This phenomenon is known as map generation and also map inference procedure. For that, we gathered big GPS data from about ten thousands of vehicles equipped with GPS receivers and circulating in Tunisia, which does not have a road map like other developing countries. We collected a big database with approximately 100 gigabytes. After preprocessing it, we were obliged to partition data in order to facilitate handling an unstructured database with a such size. In fact, we used for that K-means with its sequential mode and the parallel mode based on Mapreduce, which is one of the most famous proposed solution to analyse the rapidly growing data. The proposed parallel k-means algorithm was tested with our GPS data and the results are efficient in processing large datasets. It is a parallel data processing tool which is gathering significant importance from industry and academia especially with appearance of a new term to describe massive datasets having large-volume, high-complexity and growing data from different sources, “big data”.

The Research of Parallelization of K-Means Algorithm Based on CPU/GPU Heterogeneous Platform

Clustering analysis is widely used in data mining, e-commerce, graphic processing, bioinformation and text classification. Multicore computing based on CUDA and GPU is one of the new techniques of data processing, which became an active research direction of parallel computing in recent years. After the analysis and serial k-means algorithms, we propose a new compact parallel k-means algorithm which fit for GPU computing and present three main optimization method.The experiment results show that the algorithm is simple, fast and scalable for real-world data processing with comparison of existing other research.

Parallelization of the Algorithm K-means Applied in Image Segmentation

Algorithm k-means is useful for grouping operations; however, when is applied to large amounts of data, its computational cost is high. This research propose an optimization of k-means algorithm by using parallelization techniques and synchronization, which is applied to image segmentation. In the results obtained, the parallel k-means algorithm, improvement 50% to the algorithm sequential k-means. General Terms: speedup

Verification and Validation of MapReduce Program model for Parallel K-Means algorithm on Hadoop Cluster

Verification and Validation of MapReduce Program model for Parallel K-Means algorithm on Hadoop Cluster

Improvement and Parallelism of k-Means Clustering Algorithm

The k-means clustering algorithm is one of the most commonly used algorithms for clustering analysis. The traditional k-means algorithm is, however, inefficient while working on large numbers of data sets and improving the algorithm efficiency remains a problem. This paper focuses on the efficiency issues of cluster algorithms. A refined initial cluster centers method is designed to reduce the number of iterative procedures in the algorithm. A parallel k-means algorithm is also studied for the problem of the operation limitation of a single processor machine when given huge data sets. The analytical results demonstrate that these improvements can greatly enhance the efficiency of the k-means algorithm, i.e., allow the grouping of a large number of data sets more accurately and more quickly. The analysis has theoretical and practical importance for work on the improvement and parallelism of cluster algorithms.