A New Multiobjective Evolutionary Algorithm for Mining a Reduced Set of Interesting Positive and Negative Quantitative Association Rules

Martín, Daniel; Rosete, Alejandro; Alcalá‐Fdez, Jesús; Herrera, Francisco

doi:10.1109/tevc.2013.2285016

Cited by 91 publications

(35 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Recently, the study of the positive and negative dependencies by means of a multi-objective evolutionary algorithm has been proposed [19]. MOPNAR's fitness function consists of three measures to be optimized: the interestingness, the comprehensibility and what they named performance, a modification of the coverage.…”

Section: Related Workmentioning

confidence: 99%

Obtaining optimal quality measures for quantitative association rules

et al. 2016

View full text Add to dashboard Cite

a b s t r a c tThere exist several works in the literature in which fitness functions based on a combination of weighted measures for the discovery of association rules have been proposed. Nevertheless, some differences in the measures used to assess the quality of association rules could be obtained according to the values of the weights of the measures included in the fitness function. Therefore, user's decision is very important in order to specify the weights of the measures involved in the optimization process. This paper presents a study of well-known quality measures with regard to the weights of the measures that appear in a fitness function. In particular, the fitness function of an existing evolutionary algorithm called QARGA has been considered with the purpose of suggesting the values that should be assigned to the weights, depending on the set of measures to be optimized. As initial step, several experiments have been carried out from 35 public datasets in order to show how the weights for confidence, support, amplitude and number of attributes measures included in the fitness function have an influence on different quality measures according to several minimum support thresholds. Second, statistical tests have been conducted for evaluating when the differences in measures of the rules obtained by QARGA are significative, and thus, to provide the best weights to be considered depending on the group of measures to be optimized. Finally, the results obtained when using the recommended weights for two real-world applications related to ozone and earthquakes are reported.

show abstract

Section: Related Workmentioning

confidence: 99%

Obtaining optimal quality measures for quantitative association rules

et al. 2016

View full text Add to dashboard Cite

show abstract

“…To this [21] proposed the MOPNAR algorithm, a new multiobjective evolutionary algorithm that mines a reduced set of positive and negative quantitative association rules with low computational cost.…”

Section: Genetic Algorithmsmentioning

confidence: 99%

“…Additionally, the use of both positive and negative patterns has been addressed by different researchers [21]. Other authors have considered the importance of introducing syntax constraints into MOEA for mining patterns [20].…”

Section: Introductionmentioning

confidence: 99%

Multiobjective Approaches in Pattern Mining

Ventura

Luna

2016

Pattern Mining With Evolutionary Algorithms

View full text Add to dashboard Cite

In pattern mining, solutions tend to be evaluated according to several conflicting quality measures and, sometimes, these measures have to be optimized simultaneously. The problem of optimizing more than one objective function is known as multiobjective optimization, which together with evolutionary computation have given rise to evolutionary multiobjective optimization. This issue is well addressed in this chapter, which describes the usefulness of evolutionary multiobjective optimization in the pattern mining field. First, some fundamental concepts of evolutionary multiobjective optimization are introduced, including metrics to evaluate the quality of the solutions, and measures to be combined for the pattern mining problem. Then, different multiobjective approaches in the pattern mining field are described, which are divided into genetic algorithms, genetic programming approaches and other kind of algorithms like evolutionary computation and swarm intelligence. Finally, some useful real-world applications are analysed.

show abstract

“…[4], [5], [6], and [7]) have exponential time complexity. Due to requiring several measures to have strong rule set, thereby multi-objective evolutionary algorithms are more suitable solutions than single-objective evolutionary optimizations [8], [9].…”

Section: Introductionmentioning

confidence: 99%

“…Comprehensibility measure is used for mining simple rules. Martin, Rosete, Acala-Fedez, and Herrera used three measures: performance, lift and comprehensibility [9]. They proposed a multi-objective evolutionary algorithm (MOPNAR) in order to explore a reduced set of positive and negative association rules with low computational cost.…”

Section: Introductionmentioning

confidence: 99%

Proposing an efficient combination of interesting measures for mining association rules via NSGA-II

Rokh

Mirvaziri

Eftekhari

2014

2014 International Congress on Technology, Communication and Knowledge (ICTCK)

View full text Add to dashboard Cite

Selecting accurate and simple association rules that efficiently cover all data samples is very important in knowledge discovery. There are several measures to assess accuracy and relations in a rule. This poses a challenge for researchers to select effective measures. Combining different measures via multi-objective evolutionary algorithms is an effective method to select suitable association rules. Therefore in this paper NSGA-II algorithm is employed for rule selection via different combination of existing measures (support, certainty factor, change of support, Yao and Liu's one way support, cosine and lift) as objectives. The contributions of the paper are twofold. Firstly, some existing measures are modified. Secondly, several experiments are done to evaluate the performance of different combinations of measures through NSGA-II. The experimental results show that the combination of certainty factor and square of cosine measures are more effective in rule selection.

show abstract

A New Multiobjective Evolutionary Algorithm for Mining a Reduced Set of Interesting Positive and Negative Quantitative Association Rules

Cited by 91 publications

References 37 publications

Obtaining optimal quality measures for quantitative association rules

Obtaining optimal quality measures for quantitative association rules

Multiobjective Approaches in Pattern Mining

Proposing an efficient combination of interesting measures for mining association rules via NSGA-II

Contact Info

Product

Resources

About