Modified Soft Rough set for Multiclass Classification
Rough set theory has been applied to several domains because of its ability to handle imperfect knowledge. Most recent extension of rough set is soft rough set, where parameterized subsets of a universal set are basic building blocks for lower and upper a
- PDF / 820,151 Bytes
- 6 Pages / 439.37 x 666.142 pts Page_size
- 88 Downloads / 228 Views
Abstract Rough set theory has been applied to several domains because of its ability to handle imperfect knowledge. Most recent extension of rough set is soft rough set, where parameterized subsets of a universal set are basic building blocks for lower and upper approximations of a subset. In this paper, a new model of soft rough set, which is called modified soft rough set (MSR) where information granules are finer than soft rough sets, is applied for classification of medical data. In this paper, rough-set-based quick reduct approach is applied for selecting relevant features and MSR is applied for multiclass classification problem and the proposed work is compared with bijective soft set (BSS)-based classification, naïve Bayes, and decision table classifier algorithms based on evaluation metrics.
Keywords Soft rough set Classification Modified soft rough set Quick reduct
1 Introduction Classification and feature reduction are wide areas of research in data mining. Many practical applications of classification involve a large volume of data and/or a large number of features/attributes. Hence, it is necessary to remove irrelevant attributes and only the relevant attributes are used. The new idea is to solve multiclass classification problem with the modified soft rough set (MSR) method. S. Senthilkumar (&) H. H. Inbarani S. Udhayakumar Department of Computer Science, Periyar University, Salem 636011, India e-mail: [email protected] H. H. Inbarani e-mail: [email protected] S. Udhayakumar e-mail: [email protected]
G. S. S. Krishnan et al. (eds.), Computational Intelligence, Cyber Security and Computational Models, Advances in Intelligent Systems and Computing 246, DOI: 10.1007/978-81-322-1680-3_41, Springer India 2014
379
380
S. Senthilkumar et al.
In order to solve such a problem in medical diagnosis, attribute reduction, also called feature subset selection, is usually employed as a preprocessing step to select part of the attributes and focus the learning algorithm on relevant information. Rough set has strong ability in data processing and can extract useful rules from them [1]. The main aim of feature selection is to determine a minimal feature subset from a problem domain. A decision rule is a function, which maps an observation to an appropriate action. Deterministic rules correspond to the lower approximation, and non-deterministic rules correspond to the upper approximation [2]. Rough set theory and soft set theory are two different tools to deal with uncertainty. MSR sets satisfy all the basic properties of rough sets and soft sets. In some situations, equivalence relation cannot be defined, which is the basic requirement in rough set theory [3]. In these situations, MSR sets can help us to find approximations of subsets. Our proposed work consists of two parts: Initially, in the preprocessing stage, redundant data are removed and rules are derived from reduced data set. In this study, MSR-based classification is applied for generating decision rules from the reduced data set. Mo
Data Loading...
