Xiaochun Wang · Nova Science Publishers, Inc. eBooks · 2026
In the big data era, many modern data mining problems cannot be solved efficiently by the traditional algorithms, (i.e., specifically intractable by a single-processor computing system, either spatially prohibitive, temporally prohibitive or both). Obtaining optimal solutions in a tolerable amount of time, parallel data mining algorithms can be chosen as a suitable tool to solve the aforesaid problems. However, most up-to-date books on parallel and distributed computing techniques, such as Hadoop and Spark, are not wholly on the data mining subjects, covering only a portion of the materials in one or a few chapters and touching on the subjects but without going into it deeply, and thus incomplete and lacking systematicity, specialty and comprehensiveness. Parallel Data Mining Algorithms uniquely combines systematicity and comprehensiveness, trying not only to present as many data mining algorithms developed throughout till now as possible but also to provide their parallel solutions developed by the research and industry community currently. Instead of scratching the surface, it covers a broad range of data mining algorithms in depth, provides a comprehensive coverage of the state of the arts and advances in parallel data mining algorithm research (covering such topics as the parallel algorithm design in general and the implementations by the divide-and-conquer scheme, the MapReduce programming model and the Resilient Distributed Dataset (RDD) in specific) and makes their design and analysis accessible to all levels of readers with self-contained chapters and algorithms in pseudocode and with updated notes and bibliography to reflect developments in the field, in the hope that it will become an introduction-level parallel data mining algorithm textbook in universities as well as the standard reference for professionals.
Cited by 0
These Explanations Can, David Skillicorn · 2007
This article shows how to use such measures to assess different parallelization strategies for data-mining algorithms. We also show how the structure of some algorithms results in a doublespeedup phenomenon. A high-level cost analysis greatly simplifies the search for an effective algorithm.
Cited by 0
Oliver R. Sampson · KOPS (University of Konstanz) · 2013
With increasing availability and power of parallel computational resources, attention is drawn to the question of how best to apply those resources. Instead of simply finding the same answers more quickly, this thesis describes how parallel computational resources are used to explore disparate regions of a solution space by using diversity to steer the solution paths away from each other, thereby discouraging strictly greedy behavior. The formulation of models in a concept/solution space and its relationship to a search space are described as well as common search algorithms with heuristics for time or space computationally prohibitive searches. Measures of diversity are introduced, and the application of a beam search to the solution space for the Krimp algorithm for frequent itemset mining is described. Experimental results show that it is indeed possible to get better results on real-world datasets with these methods.
Cited by 0Open access
Wang, Y. · 2000
Data mining is the process of automatic extraction of novel, useful and understandable patterns in very large databases. High-performance, scalable, and parallel computing algorithms are crucial in data mining as datasets grow in size and complexity. Inductive logic is a research area in the intersection of machine learning and logic programming, which has been recently applied to data mining. Inductive logic studies learning from examples, within the framework provided by clausal logic. It provides a uniform and expressive means of representation: examples, background knowledge, and induced theories are all expressed in first-order logic. Such an expressive representation is computationally expensive, so it is natural to consider improving the performance of inductive logic data mining using parallelism. We present a parallelization technique for inductive logic, and implement a parallel version of a core inductive logic programming system, Progol. Performance results on seve...
Cited by 4
Bakin, Sergey, Hegland, Markus, Williams, Graham · 1999
A new parallel version of Friedman's Multivariate Adaptive Regression Splines (MARS) algorithm is discussed. By partitioning the data over the processors of a parallel computational system one achieves good parallel eciency. Instead of using truncated power basis functions of the original MARS, the new method (BMARS) utilises B-splines which improves numerical stability and reduces the computational cost of the procedure. In addition, the coecients of the basis functions of a BMARS model provide quickly accessible information about the local behaviour of the function. The algorithm has a time complexity proportional to the number of data records. The method provides a new means for the detection of areas in the space of features which are characterised by the \\interesting" patterns of response values. This is applied to searching for classes of incorrect tax returns using multiple predictor variables or features. The
Cited by 0
In Conjunction Ecml/pkdd, Michael R. Berthold, Srinivasan Parthasarathy · 2006
Cited by 0
Wenbin Fang, Ka Keung Lau, Mian Lu et al. · 2011
We introduce GPUMiner, a novel parallel data mining system that utilizes new-generation graphics processing units (GPUs). Our system relies on the massively multi-threaded SIMD (Single Instruction, Multiple-Data) architecture provided by GPUs. As specialpurpose co-processors, these processors are highly optimized for graphics rendering and rely on the CPU for data input/output as well as complex program control. Therefore, we design GPUMiner to consist of the following three components: (1) a CPU-based storage and buffer manager to handle I/O and data transfer between the CPU and the GPU, (2) a GPU-CPU co-processing parallel mining module, and (3) a GPU-based mining visualization module. We design the GPU-CPU co-processing scheme in mining depending on the complexity and inherent parallelism of individual mining algorithms. We provide the visualization module to facilitate users to observe and interact with the mining process online. We have implemented the k-means clustering and the Apriori frequent pattern mining algorithms in GPUMiner. Our preliminary results have shown significant speedups over state-of-the-art CPU implementations on a PC with a G80 GPU and a quad-core CPU. We will demonstrate the mining process through our visualization module. Code and documentation of GPUMiner are available at
Cited by 107
Xiaohong Qiu, Geoffrey Fox · 2010
As in many fields the data deluge impacts all aspects of Life Sciences from chemistry data in PubChem; genetic sequence data through health records. This data demands analysis and mining algorithms that are both high performance and robust. Further although some of the data can be usefully viewed as points in a vector space; for others it is better just to consider relationships defined just by dissimilarities between points. We develop parallel algorithms for both pairwise (no vectors) and vector based clustering using deterministic annealing for robustness and present preliminary results on medical record and gene sequence studies. We compare MPI and threading on multicore systems with up to 24 cores on an individual system and 128 cores on an 8 node cluster. In our analysis of performance and ease of programming, we note MPI is particularly effective in MapReduce scenarios and although threading is efficient at the times when MPI needs send and receive, it can have significant synchronization overhead due to runtime fluctuations on Windows.
Cited by 0
Chandrika Kamath · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 2023
A data mining system uncovers patterns, associations, anomalies and other statistically significant structures in data. Data files are read and displayed. Objects in the data files are identified. Relevant features for the objects are extracted. Patterns among the objects are recognized based upon the features. Data from the Faint Images of the Radio Sky at Twenty Centimeters (FIRST) sky survey was used to search for bent doubles. This test was conducted on data from the Very Large Array in New Mexico which seeks to locate a special type of quasar (radio-emitting stellar object) called bent doubles. The FIRST survey has generated more than 32,000 images of the sky to date. Each image is 7.1 megabytes, yielding more than 100 gigabytes of image data in the entire data set.
Cited by 0Open access
Mauro Sousa, Marta Mattoso, Nelson Ebecken · 1998
Abstract Recent years have shown the need of an automated process to discover interesting and hidden patterns in real-world databases, handling large volumes of data. This sort of process implies a lot of computational power, memory and disk I/O, which can only be provided byparallel computers. Our work contributes with a solution that integrates a machine learning algorithm, parallelism and a tightly-coupled use of a DBMS system, addressing performance problems with parallel processing and data fragmentation.
Cited by 0
SERGEY BAKlN, Markus Hegland, Graham Williams · Parallel algorithms and applications · 2000
Abstract A new parallel version of Friedman's Multivariate Adaptive Regression Splines (MARS) algorithm is discussed. By partitioning the data over the processors of a parallel computational system one achieves good parallel efficiency. Instead of using truncated power basis functions of the original MARS, the new method (BMARS) utilises B-sp!ines which improves numerical stability and reduces the computational cost of the procedure. In addition, the coefficients of the basis functions of a BMARS model provide quickly accessible information about the local behaviour of the function. The algorithm has a time complexity proportional to the number of data records. The method provides a new means for the detection of areas in the space of features which are characterised by the "interesting" patterns of response values. This is applied to searching for classes of incorrect tax returns using multiple predictor variables or features. The parallel algorithm makes it feasible to investigate very large databases, such as the taxation database.
Cited by 7
Yike Guo, Stefan M. Rüger, Janjao Sutiwaraphun et al. · 1997
In the paper, we propose a new approach to applying meta-learning concepts to parallel data mining. We name this approach Knowledge Probing where a supervised learning is organised into two learning stages. In the first learning step, a set of base classifiers are learned in parallel from a data set. In the second learning step, meta-learning is applied to induce the relationship between an attribute vector and the class predictions from all the base classifiers. This approach therefore can be used to extract knowledge from non-descriptive models derived by some data mining algorithms such as neural nets. By applying this approach to the environment where base classifiers are produced from different learning algorithms, the output of the Knowledge Probing process can be viewed as the assimilated knowledge learned from the different learning algorithms. Some initial experimental results and a technique to estimate the quality of the assimilated knowledge are presented. We believe that ...
Cited by 0
Xiaohong Qiu, Geoffrey C. Fox, Huapeng Yuan et al. · 2008
The ever increasing number of cores per chip will be accompanied by a pervasive data deluge whose size will probably increase even faster than CPU core count over the next few years. This suggests the importance of parallel data analysis and data mining applications with good multicore, cluster and grid performance. This paper considers data clustering, mixture models and dimensional reduction presenting a unified framework applicable to bioinformatics, cheminformatics and demographics. Deterministic annealing is used to lessen effect of local minima. We present performance results on 8-core systems identifying effects from cache, runtime fluctuations, synchronization and memory bandwidth. We discuss needed programming model and compare with MPI and other approaches 1.
Cited by 0
P. Pesciullesi B, L. Presti C, R. Ravazzolo B et al. · 2008
We propose the design of a data management abstraction level to implement a full set of parallel KDD applications with minimal performance overhead and greater scalability than conventional DBMS, providing a high-level parallel API to be exploited by parallel and out-of-core data mining algorithms. We describe an existing prototype and report examples and first test results with mining algorithms. 1.
Cited by 0
Massimo Coppola, P. Pesciullesi, Luigi Presti et al. · CINECA IRIS Institutial research information system (University of Pisa) · 2005
Cited by 0
P. Pesciullesi, L. Presti, R. Ravazzolo et al. · 2005
Cited by 0
Kevin A. Huck · 2009
Parallel applications running on high-end computer systems manifest a complex combination of performance phenomena, such as communication patterns, work distributions, and computational inefficiencies. Current performance tools compute results that help to describe performance behavior, as well as to understand performance problems and how they came about. Unfortunately, parallel performance tool research has been limited in its contributions to large-scale performance data management and analysis, automated performance investigation, and knowledge-based performance problem reasoning. This dissertation discusses the design of a performance analysis methodology and framework which integrates scalable data management, dimension reduction, clustering, classification and correlation analysis of individual trials of large dimensions, and comparative analysis between multiple application executions. Analysis process workflows can be captured, automating what would otherwise bev time-consuming and possibly error prone tasks. More importantly, process automation provides an extensible interface to the analysis process. The methods
Cited by 0
S. Fotis, J.A. Keane, R. I. Scott · Lecture notes in computer science · 1999
Cited by 0
Karshiev Zaynidin, Sattarov Mirzabek · International Journal of Innovative Technology and Exploring Engineering · 2019
Paper describes the technique of parallelization of data mining algorithms based on the block structure. The described technique takes into account the characteristics of the execution environment of parallel algorithms, such as the distribution of the analyzed data, as well as the features of the parallelized algorithm.
Cited by 1Open access
Zaenal Akbar, Violeta N. Ivanova, Michael R. Berthold · Lecture notes in computer science · 2012
Cited by 12