Fuzzy Rough Clustering Analysis Algorithm Based on Attribute Reduction

Decision-theoretic rough sets (DTRS) can be seen as a kind of misclassification cost-sensitive learning model. In DTRS, attribute reduction is the process of minimizing misclassification costs. However in parctice, data are not free, and there are test costs to obtain feature values of objects. Hence, the process of attribute reduction should help minimizing both of misclassification costs and test costs. In this paper, the minimal test cost attribute reduct (MTCAR) problem is defined in DTRS. The objective of attribute reduction is to minimize misclassification costs and test costs. A genetic algorithm (GA) is used to solve this problem. Experiments on UCI data sets are performed to validate the effectiveness of GA to solve MTCAR problem.KeywordsDecision-theoretic rough setsmisclassification coststest costsattribute reductcost-sensitive

Attribute reduction plays an important role in pattern recognition and machine learning, and the theory of rough sets has become a commonly used model for attribute reduction since its superiority in describing and quantifying vagueness and uncertainty. However, little attention has been paid to analyzing the importance of core attributes in different circumstances, and core attributes are generally all selected into the reduct without considering their differences. In this study, the role of core attributes is analyzed in terms of their impacts on classification ability, and the three-way partition of attributes is proposed to distinguish different core attributes and condition attributes. Then, a unified approximate attribute reduction framework based on the three-way decision is introduced to keep the quality and resulting classification performance of the reduct. Moreover, a general forward-adding and back-deleting heuristic algorithm is developed to effectively select important attributes and also eliminate unimportant core and condition attributes in the boundary region. Comprehensive comparative experiments and statistical significance analysis are conducted on UCI data sets. The experimental results show that our method achieves a better attribute reduction rate and classification performance and also verify that core attributes are not always indispensable.

Metric technique has recently been applied to solve such data mining problems as classification, clustering, feature selection, decision tree construction. In this paper, we apply metric technique to solve a attribute reduction problem of incomplete decision tables in rough set theory. We generalize Liang entropy in incomplete information systems and investigate its properties. Based on the generalized Liang entropy, we establish a metric between coverings and study its properties for attribute reduction. Consequently, we propose a metric based attribute reduction method in incomplete decision tables and perform experiments on UCI data sets. The experimental results show that metric technique is an effective method for attribute reduction in incomplete decision tables.

In view of the advantages of deep learning model in the extraction of abstract concept, a new text clustering algorithm is designed based on Deep Boltzmann Machines. Based on Replicate Softmax Model and new Deep Boltzmann Machine, energy function of this model is proposed and the detail learning algorithm is introduced.

The traditional K-means clustering algorithm is difficult to determine the cluster number, which is sensitive to the initialization of the clustering center and easy to fall into local optimum. This paper proposes a clustering algorithm based on self-organizing mapping network and weight particle swarm optimization SOM&WPSO (Self-Organization Map and Weight Particle Swarm Optimization). Firstly, the algorithm takes the competitive learning mechanism of a self-organizing mapping network to divide the data samples into coarse clusters and obtain the clustering center. Then, the obtained clustering center is used as the initialization parameter of the weight particle swarm optimization algorithm. The particle position of the WPSO algorithm is determined by the traditional clustering center is improved to the sample weight, and the cluster center is the “food” of the particle group. Each particle moves toward the nearest cluster center. Each iteration optimizes the particle position and velocity and uses K-means and K-medoids recalculates cluster centers and cluster partitions until the end of the algorithm convergence iteration. After a lot of experimental analysis on the commonly used UCI data set, this paper not only solves the shortcomings of K-means clustering algorithm, the problem of dependence of the initial clustering center, and improves the accuracy of clustering, but also avoids falling into the local optimum. The algorithm has good global convergence.

Clustering analysis is an important approach of data mining. This paper presents an ANN-based clustering analysis algorithm with dynamic data window (DDW). Comparing with k-means algorithm merged in density-based and integrated clustering analysis algorithm, the new clustering analysis algorithm based on artificial neural networks and combining with DDW has more valuable in data mining. This algorithm can immensely avoid the effect on accumulation points from boundary points, and can automatically find representative accumulation points in all kings of shapes. Furthermore its applications in CIS will be discussed in the paper. Some analysis results show the significant improvement to ship-routing design with the clustering analysis algorithm based on ANN and DDW in database of CIS.

In this paper a novel fuzzy cluster analysis algorithm based on simulated annealing (FCASA) is proposed. Instead of the perturbation criterion of the literature in which membership grades of several samples may be changed simultaneously in each perturbation or perhaps each sample has unequal probability of perturbation, the present method changes the membership grades of several samples among all samples in each perturbation cycle at first, then only the membership grade of one sample in each perturbation cycle is changed when the number of iterations at a given temperature (see the Theory section for details) equals half of the desired number (i.e. the Metropolis sampling number). The algorithm is first tested with the Fisher data set and a simulated data set, then it is used for the classification of phenolic compounds, based on their parameters of molecular connectivity index and partition coefficient calculated by the present authors, and Chinese tea samples. The results obtained from both simulated data and practical samples show that the proposed algorithm is more feasible than the general fuzzy cluster analysis (GFCA) algorithm in obtaining a global or near-global optimum. Besides, the present method based on the modified perturbation criterion can save much time in getting the optimal solution of a particular problem.

Based on multiple ethnic elements, we investigate the effectiveness of environmental art design by detailedly analyzing the mixed attribute information with real value, interval number or natural language. Considering the traditional numerical information(e.g., FCM clustering algorithm), we propose a new algorithm for clustering analysis. In this method, we first introduce the clustering analysis problem with hybrid multi index information, and put forward two theorems about determining the optimal partition and the optimal clustering center based on the hybrid multi index information, we then give the iteration steps of the FCM clustering algorithm for mixed attribute information. Based on the two optimal parameters, a case study of multi-ethnic elements is made, and the effectiveness of environmental art design is analyzed. Due to a lack of adequate characterization of single similarity measure, we employ a combined similarity clustering analysis method in this paper, i.e., make use of the combined similarity to summarize the relationship between various ethnic factors and environmental design, which greatly improves the reliability and comprehensiveness of similarity measure and, can be applying in making reasonable classification of the hybrid multi index information. This method, which is intuitive, easy to operate, and convenient for application, not only can be used in the study on environmental art design of ethnic elements, but also lays a foundation for further analysis of related issues.

A unified reduction algorithm based on invariant matrices for decision tables

Density Peaks Clustering (DPC) algorithm identifies cluster centers based on local density and relative distance, disregarding the effect of the sample's environment on its density. This makes it challenging to identify cluster centers in low-density regions. Moreover, the single-step allocation strategy of DPC algorithm lacks fault tolerance. If a sample point is incorrectly allocated, it can trigger a cascade of subsequent allocation error. To solve the aforementioned issue, a density peak clustering algorithm based on natural neighbors and diffusion allocation (NNDA-DPC) is proposed. Firstly, the concept of natural neighbors is introduced, and the local density of sample points is redefined by considering the density of sample points and the environment in which the sample points are located. This is done to reduce the influence of the sparsity of class clusters on the selection of class cluster centers. The clustering centers are automatically determined by computing the decision value. Secondly, the natural neighbor points of the selected clustering centers are assigned to the corresponding clustering centers to form the initial microclusters. Finally, the initial microclusters are formed based on the remaining sample points and the initial clusters. The final clustering result is obtained by assigning the sample points based on the correlation between the remaining sample points and the initial clusters, which helps prevent the cascading effect of assignment errors. We compared the NNDA-DPC algorithm with DPC and its enhanced versions on artificial and UCI datasets. The experimental results demonstrate that the NNDA-DPC algorithm can efficiently cluster the datasets.

A systematic study on attribute reduction with rough sets based on general binary relations

Dominance-based rough set approach (DRSA) has attracted much attention in multi-criteria decision-making problems for data with preference-ordered relations since its inception. However, in the real world, there is often the challenge of incomplete information. This paper addresses approaches to attribute reduction in an incomplete ordered decision system (IODS). Attribute evaluation is crucial in the attribute reduction process. Traditional evaluation functions primarily depend on the relationships between features, neglecting their impact on classification, and incur high computational costs when calculating lower and upper approximations. To address these issues, this paper proposes whole inter-class non-dominance as the criterion for attribute importance. First, we define the concepts of inter-class proximity and inter-class non-dominance in an IODS, with whole inter-class non-dominance serving as the evaluation index for attribute importance. Next, we propose a method for calculating inter-class non-dominance based on the expanded dominance matrix. Finally, combined with the heuristic search strategy, an attribute reduction algorithm based on inter-class non-dominance (HANDR) was designed, and experiments on various datasets from the University of California at Irvine (UCI) demonstrate that the proposed algorithm significantly improves running time and classification accuracy.

A new method based on attribute reduction, which is developed from rough set theory, is introduced to dig out main factors leading to human errors in marine accidents. The basic theory of rough set is mentioned, and a method of attribute deduction based on genetic algorithm is introduced. Experiment based on RSES, a rough set analysis tool, is implemented on a data-set which includes 30 factors which may be lead to Look-out Missing, a common human error on board. Reasonable result, an accurate set of main factors, is received after the implementation of the experiment and a conclusion that attribute reduction is a reliable and convenient way to generalize main factors lead to human errors in marine accident is made finally according to the experiment result.

In this paper an improved clustering analysis algorithm is proposed on the basis of randomly selected data in the CURE algorithm and the cluster centers setting in the K-NN algorithm. The combination of the two algorithms conquers the poor clustering accuracy in the CURE algorithm and the clustering deficiency on large data set in the K-NN algorithm. This paper first introduces the concept of clustering analysis and its real-life applications, then proceeds to describe the method of clustering analysis, and then introduces optimized CURE_KNN algorithm described in pseudo-code, at last the advantages of the algorithm are summarized.

Granular maximum decision entropy-based monotonic uncertainty measure for attribute reduction