Extraction and optimization of fuzzy association rules using multi-objective genetic algorithm

As opposed to quantitative association rule mining, fuzzy association rule mining is said to prevent the overestimation of boundary cases, as can be shown by small examples. Rule mining, however, becomes interesting in large databases, where the problem of boundary cases is less apparent and can be further suppressed by using sensible partitioning methods. A data-driven approach is used to investigate if there is a significant difference between quantitative and fuzzy association rules in large databases. The influence of the choice of a particular triangular norm in this respect is also examined.

In real-world applications, transactions are typically represented by quantitative data. Thus, fuzzy association rule mining algorithms have been proposed to handle these quantitative transactions. In addition, items generally have certain lifespans or temporal periods in which they exist in a database. Therefore, fuzzy temporal association rule mining algorithms have also been proposed in the literature. A key factor in the acquisition of fuzzy temporal association rules (FTARs) is the design of appropriate membership functions. Because current approaches have been designed to generate membership functions for mining fuzzy association rules (FARs) in market-basket analysis, in this paper, we propose a membership function tuning mechanism for a fuzzy temporal association rule mining algorithm. The proposed approach modifies an existing cluster-based method to generate unique membership functions that are specifically tailored to each item in a dataset. Two factors are utilized to decide the appropriate membership functions of each item: (1) the density similarity among intervals corresponding to the density similarity within intervals, and (2) the information closeness within an interval corresponding to the similarity in the number of data points between intervals. A parameter θ is used to indicate the relative importance of these two factors. As a result, the membership functions are generated based on the quantitative ranges of individual items, and the generated membership functions of items are different in terms of the values of each interval and the number of intervals. The generated membership functions are subsequently used in a fuzzy temporal association rule mining algorithm. Computational experiments were conducted on both a synthetic dataset and a real-world one to demonstrate the effectiveness of the proposed approach.

This paper presents a hybrid data mining mechanism to extract expert knowledge from historical data and extend expert systems' reasoning capabilities by using fuzzy neural network (FNN)-based learning & rule extraction algorithm. Our hybrid data mining mechanism is based on association rule extraction mechanism, FNN learning and fuzzy rule extraction algorithm. Most of traditional data mining mechanisms are depended ()n association rule extraction algorithm. However, the basic association rule-based data mining systems has not the learning ability. Therefore, there is a problem to extend the knowledge base adaptively. In addition, sequential patterns of association rules can`t represent the complicate fuzzy logic in real-world. To resolve these problems, we suggest the hybrid data mining mechanism based on association rule-based data mining, FNN learning and fuzzy rule extraction algorithm. Our hybrid data mining mechanism is consisted of four phases. First, we use general association rule mining mechanism to develop an initial rule base. Then, in the second phase, we adopt the FNN learning algorithm to extract the hidden relationships or patterns embedded in the historical data. Third, after the learning of FNN, the fuzzy rule extraction algorithm will be used to extract the implicit knowledge from the FNN. Fourth, we will combine the association rules (initial rule base) and fuzzy rules. Implementation results show that the hybrid data mining mechanism can reflect both association rule-based knowledge extraction and FNN-based knowledge extension.

Parallel extraction of association rules from genomics data

Rule extraction from data is one of the key technologies for solving the bottlenecks in artificial intelligence. Artificial neural networks are well suited for representing any knowledge in given data. Extraction of logical/fuzzy rules from the trained artificial neural network is of great importance to researchers in the fields of artificial intelligence and soft computing. Fuzzy rule sets are capable of approximating any nonlinear mapping relationships. Extraction of rules from data has been discussed in terms of fuzzy modeling, fuzzy clustering, and classification with fuzzy rule sets. This special issue entitled"Rule Extraction from Data" is aimed at providing the readers with good insights into the advanced studies in the field of rule extraction from data using neural networks/fuzzy rule sets. I invited seven research papers best suited for the theme of this special issue. All the papers were reviewed rigorously by two reviewers each. The first paper proposes an interesting rule extraction method from data using neural networks. Ishikawa presents a combination of learning with an immediate critic and a structural learning with forgetting. This method is capable of generating skeletal networks for logical rule extraction from data with correct and wrong answers. The proposed method is applied to rule extraction from lense data. The second paper presents a new methodology for logical rule extraction based on transformation of MLP (multilayered perceptron) to a logical network. Duck et al. applied their C-MLP2LN to the Iris benchmark classification problem as well as real-world medical data with very good results. In the third paper, Geczy and Usui propose fuzzy rule extraction from trained artificial neural networks. The proposed algorithm is implied from their theoretical study, not from heuristics. Their study enables to initially consider derivation of crisp rules from trained artificial neural network, and in case of conflict, application of fuzzy rules. The proposed algorithm is experimentally demonstrated with the Iris benchmark classification problem. The fourth paper presents a new framework for fuzzy modeling using genetic algorithm. The authors have broken new ground of fuzzy rule extraction from neural networks. For the fuzzy modeling, they have proposed a particular type of neural networks containing nodes representing membership functions. In this fourth paper, the authors discuss input variable selection for the fuzzy modeling under multiple criteria with different importance. A target system with a strong nonlinearity is used for demonstrating the proposed method. Kasabov, et al. present, in the fifth paper, a method for extraction of fuzzy rules that have different level of abstraction depending on several modifiable thresholds. Explanation quality becomes better with higher threshold values. They apply the proposed method to the Iris benchmark classification problem and to a real world problem. J. Yen and W. Gillespie address interpretability issue of Takagi-Sugeno-Kang model, one of the most popular fuzzy mdoels, in the fifth paper. They propose a new approach of fuzzy modeling that ensures not only a high approximation of the input-output relationship in the data, but also good insights about the local behavior of the model. The proposed method is applied to fuzzy modeling of sinc function and Mackey-Glass chaotic time series data. The last paper discusses fuzzy rule extraction from numerical data for high-dimensional classification problems. H.Ishibuchi, et al. have been pioneering methods for classification of data using fuzzy rules and genetic algorithm. In this last paper, they introduced a new criterion, simplicity of each rule, together with the conventional ones, compactness of rule base and classification ability, for high-dimensional problem. The Iris data is used for demonstrating their new classification method. They applied it also to wine data and credit data. I hope that the readers will be encouraged to explore the frontier to establish a new paradigm in the field of knowledge representation and rule extraction.

In business association rules being considered as important assets, play a vital role in its productivity and growth. Different business partnership share association rules in order to explore the capabilities to make effective decision for enhance-ment of business and core capabilities. The fuzzy association rule mining approach emerged out of the necessity to mine quantitative data regularly present in database. An association rule is sensitive when it violates few rules and regulation for sharing particular nature of information to third world. Like classical association rules, there is a need for some privacy measures to be taken for retaining the standards and importance of fuzzy association rules. Privacy preservation is used for valu-able information extraction and minimizing the risk of sensitive information disclosure. Our proposed model mainly focuses to secure the sensitive information revealing association rules. In our model, sensitive fuzzy association rules are secured by identifying sensitive fuzzy item to perturb fuzzified dataset. The resulting transformed FARs are analyzed to conclude/calculate the accuracy level of our model in context of newly generated fuzzy association rules, hidden rules and lost rules. Extensive experiments are carried out in order to demonstrate the results of our proposed model. Privacy preservation of maximum number of sensitive FARs by keeping minimum perturbation highlights the significance of our model.

Association rules and fuzzy association rules are vastly studied topics. Various measures for quantifying a quality of a (fuzzy) association rule were proposed in the past. In this article, we survey existing and propose some new quality measures for the whole rule bases of fuzzy association rules.

Online stores assist customers in buying the desired products online. Great competition in the e-commerce sector necessitates technology development. Many e-commerce systems not only present products but also offer similar products to increase online customer interest. Due to high product variety, analyzing products sold together similar to a recommendation system is a must. This study methodologically improves the traditional association rule mining (ARM) method by adding fuzzy set theory. Besides, it extends the ARM by considering not only items sold but also sales amounts. Fuzzy association rule mining (FARM) with the Apriori algorithm can catch the customers’ choice from historical transaction data. It discovers fuzzy association rules from an e-commerce company to display similar products to customers according to their needs in amount. The experimental result shows that the proposed FARM approach produces much information about e-commerce sales for decision-makers. Furthermore, the FARM method eliminates some traditional rules considering their sales amount and can produce some rules different from ARM.

Data mining is an important technology to reveal the patterns from crime data. Although there are many researches about this topic, less work models the relations between rates of different kinds of crime. In this paper, an algorithm based on fuzzy association rules (AR) mining is proposed to discover these relations. Two datasets, which are crimes in Chicago from 2012 to 2017 and crimes in NSW from 2008 to 2012, are used for case studies. At first, crime data is preprocessed, where every kind of crime occurring in every district during every month is counted. For a crime in a combination of district and month, the membership function, which is based on hypothesis testing, is designed to evaluate the degree to which its rate is high, normal or low, and the fuzzy transactional dataset is formed. A bridge between fuzzy transactional dataset and binary AR mining algorithm is built, so those mature tools of binary AR mining can be applied to generate fuzzy ARs. In the results of case studies, the strong relations between rates of different crime can be found. There are many interesting and surprise rules, which are worthy to be further studied by domain experts.

One of the major challenges in the content-based information retrieval and machine learning techniques is to-build-the-so-called "semantic classifier" which is able to effectively and efficiently classify semantic concepts in a large database. This paper dealt with semantic image classification based on hierarchical Fuzzy Association Rules (FARs) mining in the image database. Intuitively, an association rule is a unique and significant combination of image features and a semantic concept, which determines the degree of correlation between features and concept. The main idea behind this approach is that any image visual concept has some associated features, so that, there are strong correlations between the concepts and their corresponding features. Regardless of the semantic gap, an image concept appears when the corresponding features emerge in an image and vice versa. Specially, this paper's contribution was to propose a novel Fuzzy Association Rule for improving traditional association rules. Moreover, it was concerned with establishing a hierarchical fuzzy rule base in the training phase and setup corresponding fuzzy inference engine in order to classify images in the testing phase. The presented approach was independent from image segmentation and can be applied on multi-label images. Experimental results on a database of 6000 general-purpose images demonstrated the superiority of the proposed algorithm.

Mining association rules is one of the important tasks in the process of data mining application. In general, the input as used in the process of generating rules is taken from a certain data table by which all the corresponding values of every domain data have correlations one to each others as given in the data table. A problem arises when we need to generate the rules expressing the relationship between two or more domains that belong to several different tables in a normalized database. To overcome the problem, before generating rules it is necessary to join the participant tables into a general table by a process called denormalization. This paper shows a process of mining multidimensional fuzzy association rules from a normalized database. The process consists of two sub-process, namely sub-process of join tables (denormalization) and sub-process of mining fuzzy rules. In general, some parts of mining the fuzzy association rules has been discussed in our previous papers.

This work extends fuzzy inference-grams (fingrams) to fuzzy association rules (FAR), yielding FARFingrams. Their analysis pays attention to interpretability issues. An important open problem in association rule mining is the huge number of frequent itemsets and interesting rules to uncover and communicate to the user. FAR-Fingrams address such problem through visual analysis. They ease the selection of rules according to the user’s preferences and quality criteria. A new metric to construct fingrams is proposed, reflecting the particularities of FAR. Finally, some of the potentials of FAR-Fingrams are overviewed over a real-world problem that deals with user-preferences.

Fault diagnosis is a challenging work in large communication network. A fault will bring more than one alarms, an alarm may be caused by different faults vice versa. The relationship between faults and alarms are not accurate but fuzzy, which can not be understood using traditional Boolean logic. Also, crisp association rules use sharp partitioning and potentially introduce loss of information. Therefore, fuzzy logic is used to convert numerical attributes into linguistic terms and a fuzzy judge language set is described. We first analyze the meanings of each attribute of network alarms. Then define proper fuzzy membership functions according to the relationship between this field and the root alarm. After that we integrate MDs (membership degrees) and weights of every attribute to construct fuzzy database. At last, we come up with a new fuzzy association rules mining algorithm, which generalizes the popular hierarchy-based algorithm. Simulation experiments on a synthetic network are carried out to prove and validate the accuracy and efficiency of our algorithm with respect to correlation analysis for right positioning root alarms.

Association Rule Mining (ARM) with fuzzy logic perception smooth the progress of the easy process of mining of underlying frequent or recurrent patterns based on their own frequencies in the form of association rules from any transactional and relational datasets containing items to signify the most recent trends in the given dataset. These mined recurrent patterns or fuzzy association rules employ either for physical data analysis or also influenced to compel further mining tasks like categorization (classification) and collecting (clustering) which helps domain area experts to automate decision-making. In the concept of data mining, generally fuzzy Association Rule Mining (FARM) technique has been comprehensively adopted in transactional and relational datasets those datasets containing items who has a fewer to medium amount of attributes/dimensions. Few techniques have also adopted for high dimensional dataset also, but whether those techniques have also work for low dimensional datasets are yet to be proven out. Hence, in this paper we propose E-FAR-HD algorithm which is an enhanced version of FAR-HD algorithm that designed exclusively for large or high-dimensional datasets. We have designed this EFAR-HD algorithm that increases the accuracy of FAR-HD algorithm on the smaller datasets and remove the chances of misses when FAR-HD has tested on smaller datasets such as contact lens or patient dataset.

Association rules are frequently used for describing association (i.e., co-occurrence) among attribute values in the field of data mining. When an attribute is continuous (i.e., real-valued) such as height, length and weight, its domain is usually discretized into several intervals. Fuzzy rules are recognized as a convenient tool for handling continuous attributes in a human understandable manner. When we use fuzzy rules, the domain of each continuous attribute is discretized into several fuzzy sets. A linguistic label is usually associated with each fuzzy set especially when linguistic interpretations of fuzzy rules are required. In this paper, we first fuzzify the concept of association rules. That is, we show fuzzy versions of two measures (i.e., confidence and support) that are used for evaluating each association rule in the field of data mining. Then we illustrate these two measures of fuzzy rules for function approximation and pattern classification problems. Finally we examine the relation between the classification performance of fuzzy rules and the definition of their certainty grades through computer simulations. Simulation results show that the direct use of confidence as a certainty grade is not always appropriate from the viewpoint of classification performance.