An Accelerator for Rule Induction in Fuzzy Rough Theory

2007 IEEE International Conference on Intelligent Computer Communication and Processing, 2007

rule determined by a reduct contains a minimal number Abstract of attributes in its condition part, so that the set of classification rules defines the classes of decision. Rough Set Theory is successfully used within data Paragraph 2 describes both the fundamental mining for prediction, classification [1][5][1 ][12][13], definitions of Rough Set Theory and the classification discovery of associations [7] and reduction of definitions from the viewpoint of Rough Set Theory. attributes [14][17]. Having as motto "Let data speak Paragraph 3 gives a short description of the Reduct for themselves", we can state that this approach is not Equivalent Feature Selection method. Paragraph 4 1-4244-1491-1/07/$25.OO ©32007 IEEE 9

2018

The design of interpretable classifiers is a major goal in machine learning since many applications rely on a complete understanding of the learning model decisions. Among all interpretable models available in the literature, Rough sets based models have become popular given the capability of rough sets to model imprecise data. Despite its success, some of the most used rough sets models are designed to work for categorical input data. Since this design choice may severely limit the application of such models in real world problems, in this paper, we aim at showing some strategies to enable rough sets based models to work with numerical datasets.

2013

In this paper we propose a fuzzy rule generation approach based on granular computing using rough mereology (FRGAGCRM). The proposed system works in two phases. In the first phase, the pre-processing phase which use fuzzification methodology which map the numeric dataset into categorical dataset according to membership function described in this paper. In other hand, the second phase consists of the rough mereology phase and rule generation phase. The rules are generated by rough mereology and rough inclusion techniques. The experimental results using Irvine, ML Repository dataset show that FRGAGCRM is better than RGAGCRM, ID3, and RGAGC. The performance measure of FRGAGCRM is compared with the performance of RGAGCRM, ID3, and RGAGC in same dataset. The result obtained empirically demonstrates that the proposed FRGAGCRM performs superior.

International Journal of Emerging Scientific Research, 2022

In this era of big data in which, nearly every twenty months, the magnitude of data and information from various sources in the world gets doubled [1]. Due to this data explosion, there is a need for this immense data to be put to use judiciously, which is why a lot of research scholars are working tirelessly to come up with sophisticated ways for efficient information retrieval and processing. Machine learning is a field of artificial intelligence that deals with building applications, which learn from data and enhance their accuracy over time without being programmed to do so. The main function of a machine learning algorithm is to transform datasets into useful models depending on the type of data and problem domain. In most cases, raw data is always very noisy, making it difficult to use it directly by machine learning algorithms. Hence, modelling requires the selection of input parameter vectors and the corresponding output vectors that sufficiently characterize the component to be modelled [2].

International Journal of Computer Applications, 2012

The Rough Set (RS) theory can be considered as a tool to reduce the input dimensionality and to deal with vagueness and uncertainty in datasets. Over the years, there has been a rapid growth in interest in rough set theory and its applications in artificial intelligence and cognitive sciences, especially in research areas such as machine learning, intelligent systems, inductive reasoning, pattern recognition, data preprocessing, knowledge discovery, decision analysis, and expert systems. This paper discusses the basic concepts of rough set theory and point out some rough set-based research directions and applications. The discussion also includes a review of rough set theory in various machine learning techniques like clustering, feature selection and rule induction.

Granular computing has been applied in many fields to solve problems and describe spaces at different granularity and hierarchies. This paper proposes a rule generation approach based on granular computing using rough mereology. The proposed approach generates a single rule granule from granular space in each step instead of selecting a suitable attribute according to some measure and improve the performance of rule granule generation. Rules are generated by using voting strategy based on weights computed by means of rough inclusions induced from residual implications of continuous t-norm. Experimental results obtained show that the overall accuracy rate average of dataset is 64.7%, which is higher than other approaches applying decision tree algorithms. Also, the resulted average rule length is 2.25, which is less than average rule length generated using decision tree algorithms.

Lecture Notes in Computer Science, 2001

The problem of uncertain and/or incomplete information in information tables is addressed in the paper, mainly as far as the induction of classification rules is concerned. Two rule induction algorithms are introduced, discussed and tested on a number of benchmark data sets. The results obtained are promising, but further investigation can be undertaken in order to obtain more effective classification strategies.

Lecture Notes in Computer Science, 2007

The paradigm of Granular Computing has quite recently emerged as an area of research on its own; in particular, t is pursued within rough set theory initiated by Zdzis law Pawlak. Granules of knowledge consist of entities with a similar in a sense information content. An idea of a granular counterpart to a decision/information system has been put forth, along with its consequence in the form of the hypothesis that various operators, aimed at dealing with information, should factorize sufficiently faithfully through granular structures [7], [8]. Most important such operators are algorithms for inducing classifiers. We show results of testing few well-known algorithms for classifier induction on well-used data sets from Irvine Repository in order to verify the hypothesis. The results confirm the hypothesis in case of selected representative algorithms and open a new prospective area of research.

In this paper, we introduce a new automated Fuzzy-Based Rough Decision Model algorithm. Our Algorithm consists of three phases are: () automatic attributes fuzzification, () eliminate redundant attributes using rough set theory, () Generating Fuzzy rough rules then calculate automatically fitness value (Confidence) and support for each rule. In phase one, the user input the number of fuzzy sets of each attributes, our algorithm determine the maximum and minimum values of each attribute that define and calculates automatically the width (∆) which divides the universe of discourse of each attribute into " n " intervals according to the number of fuzzy sets, also the algorithm calculates automatically the width (δ i) according to the width (∆). In phase two, we use the rough set techniques to reduce the number of attributes that comes from phase one and produce fuzzy-rough rules. In phase three, our algorithm automatically calculates the confidence (weight or fitness value) and support of each fuzzy rough rule then it calculates the total weight or fitness value of all linguistic rules. The result of fitness value of our algorithm that applied on dataset in (X.Hu, T.Lin, and J.Han.) [ ] is. Rough sets theory was first introduced by Pawlak in the s [ , , , ] and it has been applied in many applications such as machine learning, knowledge discovery, and expert systems since then. It provides powerful tools for data analysis and data mining from imprecise and ambiguous data. Many rough sets models have been developed in the rough set community in the last decades [ , , ]. Applying the traditional rough set models in large data sets in data mining applications has shown that one of the strong drawbacks of the classical rough set theory assumes that all attributes values are discrete. In real life datasets values of attributes could be both of symbolic and real-valued. Therefore, the traditional rough set theory will have difficulty in handling such values. There is a need for some methods which have the capability of utilizing set approximations and attributes reduction for real-valued attributed dataset. This can be done by combining (integrate) fuzzy sets and rough sets in a Fuzzy-Based Rough Model [ , , , , , ]. Another major drawback of traditional Fuzzy-Based Rough Model is that the linguistic values (fuzzy sets) for numeric values of each attribute should determining by the membership functions of these linguistic terms which, every element in the universe of discourse is a member of the fuzzy set with some grade (degree of membership functions). Therefore, the user should define the parameters of membership functions of these linguistic values from his view which is different from one user to another. Therefore, we propose a new automated Fuzzy-Based Rough Decision Model algorithm that can define the parameters of membership functions of these linguistic values automatically that the user determine only the number of fuzzy sets (linguistic values) then the maximum and minimum values of each attribute are determined automatically then the algorithm calculates the width (∆) that divides the universe of discourse " u " of each attribute into " n " intervals according to the number of fuzzy sets then the algorithm calculates automatically the width (δi) according to the width (∆). Another strong drawback of the traditional rough set theory is the inefficiency of rough set methods and algorithms of computing the core attributes and reduct and identifying the dispensable attributes, which limits the suitability of the traditional rough set model in data mining applications. Further investigation of the problem reveals that most existing rough set models [ , , , ] do not integrate with the relational database systems and a lot of computational intensive operations are performed in flat files rather than utilizing the high performance database set operations. Moreover, not much attention and attempt have been paid to design new rough sets model by effectively combining database technologies to generate the core attributes and reducts so as to make their computations efficient and scalable in the large data set. To overcome this problem, a New Rough Sets Model Based on Database Systems has been introduced [ , ] for this purpose to redefine some concepts of rough set theory such as core attributes and reducts by using relational algebra so that the computation of core attributes and reducts can be performed with very efficient set-oriented database operations, such as Cardinality to denote the count and Projection. The paper is organized as follows: We give an overview of the rough set theory based on the model proposed by Pawlak [ , ] with some examples in Section. In Section , we give an