Efficient Discovery of Association Rules and Frequent Itemsets through Sampling with Tight Performance Guarantees

被引:38
|
作者
Riondato, Matteo [1 ]
Upfal, Eli [1 ]
机构
[1] Brown Univ, Dept Comp Sci, Providence, RI 02912 USA
基金
美国国家科学基金会;
关键词
Association rules; data mining; frequent itemsets; sampling; VC-dimension; VC-DIMENSION; ALGORITHM;
D O I
10.1145/2629586
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The tasks of extracting (top-K) Frequent Itemsets (FIs) and Association Rules (ARs) are fundamental primitives in data mining and database applications. Exact algorithms for these problems exist and are widely used, but their running time is hindered by the need of scanning the entire dataset, possibly multiple times. High-quality approximations of FIs and ARs are sufficient for most practical uses. Sampling techniques can be used for fast discovery of approximate solutions, but works exploring this technique did not provide satisfactory performance guarantees on the quality of the approximation due to the difficulty of bounding the probability of under- or oversampling any one of an unknown number of frequent itemsets. We circumvent this issue by applying the statistical concept of Vapnik-Chervonenkis (VC) dimension to develop a novel technique for providing tight bounds on the sample size that guarantees approximation of the (top-K) FIs and ARs within user-specified parameters. The resulting sample size is linearly dependent on the VC-dimension of a range space associated with the dataset. We analyze the VC-dimension of this range space and show that it is upper bounded by an easy-to-compute characteristic quantity of the dataset, the d-index, namely, the maximum integer d such that the dataset contains at least d transactions of length at least d such that no one of them is a superset of or equal to another. We show that this bound is tight for a large class of datasets. The resulting sample size is a significant improvement over previous known results. We present an extensive experimental evaluation of our technique on real and artificial datasets, demonstrating the practicality of our methods, and showing that they achieve even higher quality approximations than what is guaranteed by the analysis.
引用
收藏
页数:32
相关论文
共 28 条
  • [1] Efficient frequent itemsets mining through sampling and information granulation
    Zhang, Zhongjie
    Pedrycz, Witold
    Huang, Jian
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2017, 65 : 119 - 136
  • [2] An Efficient Algorithm for Mining Association Rules using Confident Frequent Itemsets
    Al-Maqaleh, Basheer Mohamad
    Shaab, Saleem Khalid
    2013 THIRD INTERNATIONAL CONFERENCE ON ADVANCED COMPUTING & COMMUNICATION TECHNOLOGIES (ACCT 2013), 2013, : 90 - 94
  • [3] Mining Traditional Association Rules using Frequent Itemsets Lattice
    Vo, Bay
    Le, Bac
    CIE: 2009 INTERNATIONAL CONFERENCE ON COMPUTERS AND INDUSTRIAL ENGINEERING, VOLS 1-3, 2009, : 1401 - +
  • [4] Mining Frequent Itemsets through Progressive Sampling with Rademacher Averages
    Riondato, Matteo
    Upfal, Eli
    KDD'15: PROCEEDINGS OF THE 21ST ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 2015, : 1005 - 1014
  • [5] DoS detections based on association rules and frequent itemsets
    George S Oreku
    Fredrick JMtenzi
    李建中
    Journal of Harbin Institute of Technology, 2008, (02) : 283 - 289
  • [6] EXTRACT: New extraction algorithm of association rules from frequent itemsets
    Feddaoui, Ilhem
    Felhi, Faical
    Akaichi, Jalel
    PROCEEDINGS OF THE 2016 IEEE/ACM INTERNATIONAL CONFERENCE ON ADVANCES IN SOCIAL NETWORKS ANALYSIS AND MINING ASONAM 2016, 2016, : 752 - 756
  • [7] Mining Association Rules Directly using ACO without Generating Frequent Itemsets
    Manju
    Kant, Chander
    2015 INTERNATIONAL CONFERENCE ON ENERGY SYSTEMS AND APPLICATIONS, 2015, : 390 - 395
  • [8] Mining association rules for classification using frequent generator itemsets in arules package
    Ledmi, Makhlouf
    Souidi, Mohammed El Habib
    Hahsler, Michael
    Ledmi, Abdeldjalil
    Kara-Mohamed, Chafia
    INTERNATIONAL JOURNAL OF DATA MINING MODELLING AND MANAGEMENT, 2023, 15 (02) : 203 - 221
  • [9] Effective algorithm of mining frequent itemsets for association-rules
    Liu, PQ
    Li, ZZ
    Zhao, YL
    PROCEEDINGS OF THE 2004 INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND CYBERNETICS, VOLS 1-7, 2004, : 1447 - 1451
  • [10] An Interestingness Measure and Computation Method of Association Rules Based on Frequent Itemsets Relatedness
    Chen, Xiang
    Zhou, Xuefeng
    Zhang, Yong
    FRONTIERS OF GREEN BUILDING, MATERIALS AND CIVIL ENGINEERING, PTS 1-8, 2011, 71-78 : 4039 - +