Abstract:A hybrid information system (HIS) refers to an information system (IS) including categorical attributes and numerical attributes. For an HIS, it is common to use neighborhood rough set model (NRS-model) to process different types of data. However, NRS-model applies the same neighborhood radius to deal with numerical data, which may lead to a decrease in classification ability of data. Fuzzy evidence theory is an important uncertainty reasoning method. It has been widely used in medical diagnosis, decision analysis and information fusion and other fields by virtue of its strengths of depicting uncertainty or unknown information. In this paper, fuzzy rough set model (FRS-model) based on statistical distribution of data is proposed to handle hybrid data and this model is combined with fuzzy evidence theory to design attribute reduction algorithms. Firstly, according to the statistical distribution of hybrid data, fuzzy similarity relation for each attribute is constructed and then FRS-model is established on the basis of the constructed fuzzy similarity relation. The advantage of this model lies in using different fuzzy similarity radii to construct fuzzy similarity relations, and improving the classification ability of data. After that, fuzzy belief map and fuzzy plausibility map are defined by means of fuzzy evidence theory. Subsequently, the above two maps are used to design attribute reduction algorithms. Furthermore, an overlap degree function is proposed to pre-sort attributes when designing algorithms, which can greatly reduce the times of searching for an attribute subset and ensure that the attributes with high separability are selected first. Finally, the experimental results indicate that the designed algorithms perform well in classification tasks and outlier detection tasks compared to some existing algorithms.

Hybrid Attribute Reduction for Classification Based on A Fuzzy Rough Set Technique

Face Recognition Based on Fuzzy Rough Set Reduction

Attribute Reduction with Fuzzy Rough Self-Information Measures

Novel algorithm for attribute reduction based on mutual-information gain ratio

Attribute reduction for hybrid data based on statistical distribution of data and fuzzy evidence theory

Fuzzy Rough Attribute Reduction Based on Fuzzy Implication Granularity Information

Attribute Reduction Based on the Minimum Hybrid Entropy

Attribute Reduction for Hierarchical Classification Based on Improved Fuzzy Rough Set

New Attribute Reduction on Information Entropy

Reduct Algorithm Based on Information Entropy and Rough Set Theory

Attribute reduction based on fusion information entropy

Similarity-based Attribute Reduction in Rough Set Theory: a Clustering Perspective

Attribute Reduction Based on Improved Information Entropy

Approximate Accuracy Approaches to Attribute Reduction for Information Systems

A novel attribute reduction method based on intuitionistic fuzzy three-way cognitive clustering

Fuzzy Decision-theoretic Rough Set Model and Its Attribute Reduction

Important Measure of Attributes in Fuzzy Information Systems

A new approach for attribute reduction from decision table based on intuitionistic fuzzy topology

A novel attribute reduction approach using coverage-credibility-based rough decision entropy for interval-valued data

New uncertainty measurement for hybrid data and its application in attribute reduction

Attribute Reduction Method Based on Inter-Class Separability*