文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 高等教育 > 管理学 > Method of indicator reduction based on cluster analysis and rough set theory and its application(1)1
侵权投诉

Method of indicator reduction based on cluster analysis and rough set theory and its application(1)1

discuss the attribute reduction of the interval-valued fuzzy information systems, and present the methods of the knowledge discovery for the interval-valued fuzzy information systems with an example.

Qinghua Hu et al (2008) generalize Pawlak’s rough set model into neighborhood rough set model and k-nearest-neighbor rough set model to deal with mix features. They define significance of mixed features and construct a greedy attribute reduction algorithm. Compared with other algorithm in terms of the number of selected features and classification performance, the greedy attribute reduction algorithm is effective.

Changzhong Wang et al (2008) provide a systematic study on attribute reduction with rough sets based on general binary relations. They define three systems and their attribute reductions. And compute all the reducts based on discernibility matrix.

Chris Cornelis et al (2010) present a generalization of the classical rough set framework within the context of fuzzy rough set theory for data-based attribute selection and reduction using fuzzy tolerance relations. Experimental results demonstrate the potential of fuzzy decision reducts to discover shorter attribute subsets, leading to decision models with a better coverage and with comparable, or even higher accuracy.

Yu Cao et al (2011) apply rough set theory to reduce attribute. Influence of noise data and redundant data are eliminated in this way. And the process lays the foundation for neural networks training.

Meanwhile, there are still other methods besides rough set theory such as genetic algorithms and support vector machine.

Seral Öz en, Salih Güne (2009) use genetic algorithms to determine the weights of attributes and then apply these weights in previously developed Artificial Immune System (AWAIS). After evaluating the performance of new configuration (GA-AWAIS) on two medical datasets and comparing it with AWAIS, they obtain good classification accuracy with respect to both AWAIS and other common classifiers in literature.

K.C. Tan et al (2009) integrated genetic algorithms (GAs) and support vector machines (SVMs) to select attribution. They use the GA component to search for the best attribute set and the SVM to classify the patterns in the reduced datasets, corresponding to the attribute subsets represented by the GA chromosomes. This method is proved validated using datasets obtained from the UCI machine learning repository

To be concluded, rough set theory is an important method for attribute reduction.

Compared with the second way, researches on reducing by indicators’ correlation are much fewer. Methods applied are usually normality test, parametric and non-parametric test, analysis of multicollinearity, or correlation analysis.

On one hand, reducing attribute by importance can’t guarantee that the final indicators in models are not correlative, which may result in inefficiency of models. On the other hand, reducing attribute by correlation may end up with too many unimportant indicators in models, which can lengthen the time needed to compute and get results while reducing reasonability of models. Some researchers once tried to integrate the two ways by firstly classify indicators into different layers and then use rough set theory to reduce attributes, or to firstly reduce redundant indicators through rough set theory and reduce dimensions and correlations between indicators by principal component analysis.

In conclusion, most researchers reduce attributes from the aspect of indicators’ importance to

上一页第2页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email