Association rule mining using binary particle swarm optimization

In this paper, we developed a binary particle swarm optimization (BPSO) based association rule miner. Our BPSO based association rule miner generates the association rules from the transactional database by formulating a combinatorial global optimization problem, without specifying the minimum support and minimum confidence unlike the a priori algorithm. Our algorithm generates the best M rules from the given database, where M is a given number. The quality of the rule is measured by a fitness function defined as the product of support and confidence. The effectiveness of our algorithm is tested on a real life bank dataset from commercial bank in India and three transactional datasets viz. books database, food items dataset and dataset of the general store taken from literature. Based on the results, we infer that our algorithm can be used as an alternative to the a priori algorithm and the FP-growth algorithm.     

基于命题逻辑的关联规则挖掘算法L-Eclat

关联规则挖掘是数据挖掘领域非常重要的课题,在很多领域被广泛应用。关联规则挖掘算法都需要设置最小支持度和最小置信度。很多国内外学者研究的挖掘算法在这两方面都存在着一些问题,不仅需要大量的领域知识来设置合适的最小支持度,而且其结果集庞大、用户不容易理解。针对关联规则挖掘算法存在的问题,将命题逻辑融合到关联规则算法Eclat中,设计出了基于命题逻辑思想的挖掘算法L-Eclat。实验结果表明,L-Eclat算法压缩了挖掘的规则集,减小了算法的时间消耗,且即使是非常小的支持度也可以得到高质量的关联规则,这在一定程度上解决了支持度设置的问题。     

协同进化算法在关联规则挖掘中的应用

文中采用了一种协同进化算法,分别利用改进的遗传算法和粒子群算法对两个种群同时进行迭代,并在种群之间引入一种信息交互机制,使两个种群协同进化。文中最后通过实验对该协同进化算法、传统的遗传算法以及粒子群算法应用于关联规则挖掘时的性能进行比较,证明了该协同进化算法在可接受的时间复杂度前提下,不仅继承了传统遗传算法挖掘关联规则时无须产生规模庞大的候选项集和有效减少扫描数据库次数的优点,更弥补了其容易早熟收敛的缺陷,从而能高效地搜索出数据库中高质量的关联规则,这点在其应用于高维数据集时尤为显著。     

粒子群优化算法在关联规则挖掘中的研究综述

关联规则挖掘是数据挖掘中的重要领域,考虑到当前数据的大规模、高维度、模态多样及类型复杂等特性,传统关联规则挖掘算法已无法适应大数据的需求,粒子群优化算法作为一种高效的智能优化算法,为其提供了一种全新的解决方案,近年来被广泛应用于该领域.首先对粒子群优化算法的基本原理及关联规则的基本概念进行了详细介绍,回顾了粒子群优化算...     

基于关联规则的搭配进货系统的研究与实现

为了解决商品进货无关联的现状,找到商品间的关联规则,更好地进行商品的搭配进货,从而提高进货效率,文中引入了关联规则的思想,并利用规则进行了商品关联规则的挖掘。在分析了关联规则挖掘的算法后,将其应用到超市商品数据库中,利用关联规则挖掘出大量数据中项集即商品之间的相互关联,并抽取出有价值的商品关联规则,利用支持度和平衡度这两个度量概念,优化出强规则集,并用这一思想成功设计了PLM即产品全生命周期管理中的搭配进货系统。     

改进的PSOGM算法在动态关联规则挖掘中的应用

针对动态关联规则挖掘中支持度向量和置信度向量变化趋势的分析和预测,提出一种改进的粒子群优化的灰色模型应用在动态关联规则挖掘中。由于灰色模型在引入背景值后导致在非平稳序列中的预测精度下降,因此有必要引入参数进行修正,通过在粒子群优化算法中引入二次搜索机制,优化求解灰色模型不同时刻的背景值,从而提高粒子群算法的局部搜索能力,进而提高灰色模型的预测精度。通过在Matlab平台上进行实验仿真,数据集采用超市购物数据,结果表明该方法比原始灰色模型、遗传算法优化的灰色模型和标准的粒子群优化的灰色模型具有更高的预测精度。     

Rough particle swarm optimization and its applications in data mining

This paper proposes a novel particle swarm optimization algorithm, rough particle swarm optimization algorithm (RPSOA), based on the notion of rough patterns that use rough values defined with upper and lower intervals that represent a range or set of values. In this paper, various operators and evaluation measures that can be used in RPSOA have been described and efficiently utilized in data mining applications, especially in automatic mining of numeric association rules which is a hard problem.     

Multi objective association rule mining with genetic algorithm without specifying minimum support and minimum confidence

Multi objective processing can be leveraged for mining the association rules. This paper discusses the application of multi objective genetic algorithm to association rule mining. We focus our attention especially on association rule mining. This paper proposes a method based on genetic algorithm without taking the minimum support and confidence into account. In order to improve algorithm efficiency, we apply the FP-tree algorithm. Our method extracts the best rules that have best correlation between support and confidence. The operators of our method are flexible for changing the fitness. Unlike the Apriori-based algorithm, it does not depend on support. Experimental study shows that our technique outperforms the traditional methods.     

正负关联规则两级置信度阈值设置方法

针对传统正负关联规则置信度阈值设置方法难以控制低可信度规则数量和易遗漏有趣规则的问题,提出了一个结合项集相关性的两级置信度阈值设置方法（PNMC-TWO）。首先,基于规则的无矛盾性、有效性和有趣性考虑,以相关度-支持度-置信度为框架,从规则置信度与项集支持度的计算关系出发,系统地分析了正负关联规则置信度取值随规则的项集支持度大小变化的规律;然后,与实际挖掘中用户对高可信度且有趣的规则需求相结合,提出了一个新的设置模型,避免了传统方法设置阈值时的盲目性和随意性;最后,从规则数量和规则质量两方面对所提方法与原双阈值法进行了实验对比。实验结果表明,所提方法不仅可以更好地确保提取出的关联规则有效和有趣,还可以显著地降低可信度低的关联规则数量。     

Deriving support threshold values and membership functions using the multiple-level cluster-based master–slave IFG approach

Today, development of e-commerce has provided many transaction databases with useful information for investigators exploring dependencies among the items. In data mining, the dependencies among different items can be shown using an association rule. The new fuzzy-genetic (FG) approach is designed to mine fuzzy association rules from a quantitative transaction database. Three important advantages are associated with using the FG approach: (1) the association rules can be extracted from the transaction database with a quantitative value; (2) extracting proper membership functions and support threshold values with the genetic algorithm will exert a positive effect on the mining process results; (3) expressing the association rules in a fuzzy representation is more understandable for humans. In this paper, we design a comprehensive and fast algorithm that mines level-crossing fuzzy association rules on multiple concept levels with learning support threshold values and membership functions using the cluster-based master–slave integrated FG approach. Mining the fuzzy association rules on multiple concept levels helps find more important, useful, accurate, and practical information.     

一种基于矩阵的强关联规则生成算法*

针对Apriori算法扫描数据库的I/O代价和候选项集数目较多等问题,提出一种基于矩阵的强关联规则生成算法,算法通过将事务数据库转化为0-1矩阵后对项集按照支持度计数非递减顺序排列,从而减少候选项集的产生,同时实现置信度的高效计算。通过对实例和大数据量数据库的分析表明,该方法是有效的。     

一种改进的正负关联规则挖掘算法

针对传统正负关联规则挖掘算法需要多次扫描数据库并且生成大量候选频繁项集的问题,在对比目前相关研究成果的基础上,提出了一种改进的正负关联规则挖掘算法,它通过两次数据扫描完成对正负关联规则的挖掘,对最大频繁项集的挖掘算法做了改进,有效提高了算法效率,同时对置信度标准做了改进。基于某真实事务集的实验表明,算法提高了规则挖掘的质量和有效性。     

一种新的关联规则挖掘算法研究

:通过分析数据关联的特点和已有的关联规则挖掘算法 ,在定量描述的准确性和算法高效性方面作了进一步研究 ,提出了更准确的支持度和置信度定量描述方法和关联关系强弱的定量描述方法。同时 ,改进了 FP-growth挖掘算法 ,并应用于中医舌诊临床病例数据库挖掘实验中 ,可成功准确地提取中医舌诊诊断规则。测试结果表明该算法速度快、准确度高。     

Query-based association rule mining supporting user perspective

Two parameters, namely support and confidence, in association rule mining, are used to arrange association rules in either increasing or decreasing order. These two parameters are assigned values by counting the number of transactions satisfying the rule without considering user perspective. Hence, an association rule, with low values of support and confidence, but meaningful to the user, does not receive the same importance as is perceived by the user. Reflecting user perspective is of paramount importance in light of improving user satisfaction for a given recommendation system. In this paper, we propose a model and an algorithm to extract association rules, meaningful to a user, with an ad-hoc support and confidence by allowing the user to specify the importance of each transaction. In addition, we apply the characteristics of a concept lattice, a core data structure of Formal Concept Analysis (FCA) to reflect subsumption relation of association rules when assigning the priority to each rule. Finally, we describe experiment results to verify the potential and efficiency of the proposed method.     

Efficient algorithms for distortion and blocking techniques in association rule hiding

Data mining provides the opportunity to extract useful information from large databases. Various techniques have been proposed in this context in order to extract this information in the most efficient way. However, efficiency is not our only concern in this study. The security and privacy issues over the extracted knowledge must be seriously considered as well. By taking this into consideration, we study the procedure of hiding sensitive association rules in binary data sets by blocking some data values and we present an algorithm for solving this problem. We also provide a fuzzification of the support and the confidence of an association rule in order to accommodate for the existence of blocked/unknown values. In addition, we quantitatively compare the proposed algorithm with other already published algorithms by running experiments on binary data sets, and we also qualitatively compare the efficiency of the proposed algorithm in hiding association rules. We utilize the notion of border rules, by putting weights in each rule, and we use effective data structures for the representation of the rules so as (a) to minimize the side effects created by the hiding process and (b) to speed up the selection of the victim transactions. Finally, we study the overall security of the modified database, using the C4.5 decision tree algorithm of the WEKA data mining tool, and we discuss the advantages and the limitations of blocking.     

一种基于倒排索引树的增量更新关联挖掘算法

增量更新关联规则挖掘主要解决事务数据库中交易记录不断更新和最小支持度发生变化时关联规则的维护问题。针对目前诸多增量更新关联规则挖掘算法存在效率低、计算成本高、规则难以维护等问题,提出一种基于倒排索引树的增量更新关联挖掘算法。该算法有效地将倒排索引技术与树型结构相结合,使得交易数据库中的数据不断更新和最小支持度随应用环境不同而不断改变时,以实现无需扫描原始交易数据库和不产生候选项集的情况下生成频繁项集。实验结果表明,该算法只需占用较小的存储空间、且检索项集的效率较高,能高效地解决增量更新关联规则难以维护的问题。     

一种改进的最大频繁集发现算法

关联规则是数据挖掘的主要技术,而最大频繁集是关联规则挖掘的核心。关联规则发现的准确性与效率的好坏直接决定了发现的知识规则是否适用。阐述了关联规则、频繁集和频繁超集的定义,分析了现有关联规则算法的思想及其不足,然后在概率的基础上引入了期望长,提出了ELMFI算法,最后用实例进行仿真实验并做了比较分析。该算法直接产生期望长度的候选项集并进行验算,试验结果验证了其可行性,发现效率有所提高,能节约大量的系统空间和运算时间。     

高置信度关联规则的挖掘

传统的关联规则和基于效用的关联规则,会忽略一些支持度或效用值不高、置信度（又称可信度）却非常高的规则,这些置信度很高的规则能帮助人们满足规避风险、提高成功率的期望。为挖掘这些低支持度（或效用值）、高置信度的规则,提出了HCARM算法。HCARM采用了划分的方法来处理大数据集,利用新的剪枝策略压缩搜索空间。同时,通过设定长度阈值minlen,使HCARM适合长模式挖掘。实验结果表明,该方法对高置信度长模式有效。     

一种基于向量的概率加权关联规则挖掘算法

关联规则挖掘是数据挖掘领域中最活跃的一个分支。目前提出的许多关联规则挖掘算法需要多次扫描数据库并产生大量候选项集,影响了挖掘效率。针对加权关联规则挖掘算法中多次扫描数据库影响算法性能的问题,对其进行了优化,采取了以空间换时间的思路,提出一种基于向量的概率加权关联规则挖掘算法。以求概率的方式设置项目属性的权值,通过矩阵向量存储结构保存事务记录,只需扫描一次数据库,并且采用不同的剪枝策略及加权支持度和置信度的计算方式。使用数据实例进行模拟实验,结果表明此算法明显提高了挖掘效率。     

基于垂直数据分布的大型稠密数据库快速关联规则挖掘算法

为进一步解决对大型事务数据库进行关联规则挖掘时产生的CPU时间开销大和I/O操作频繁的问题,给出了一种基于垂直数据分布的改进关联规则挖掘算法,称为VARMLDb算法。该算法首先有效地把数据库分为内存可以满足要求的若干划分,然后结合有向无环图和垂直数据形式diffse、差集来存储和计算频繁项集,极大地减少了存储中间结果所需的内存大小,解决了传统垂直数据挖掘算法对稠密数据库挖掘效率低下的问题,使该算法可有效地适用于大型稠密数据库的关联规则挖掘。整个算法吸取CARMA算法的优势,只需扫描两次数据库便可完成挖掘过程。实验结果表明该算法是正确的,在大型稠密数据库中,VARMLDb算法具有较高的执行效率。