Mustafa Mat Deris

A rough set approach for selecting clustering attribute

A few of clustering techniques for categorical data exist to group objects having similar characteristics. Some are able to handle uncertainty in the clustering process while others have stability issues. However, the performance of these techniques is an issue due to low accuracy and high computational complexity. This paper proposes a new technique called maximum dependency attributes (MDA) for selecting clustering attribute. The proposed approach is based on rough set theory by taking into account the dependency of attributes of the database. We analyze and compare the performance of MDA technique with the bi-clustering, total roughness (TR) and min-min roughness (MMR) techniques based on four test cases. The results establish the better performance of the proposed approach.

A Comparative Study for Outlier Detection Techniques in Data Mining

Existing studies in data mining mostly focus on finding patterns in large datasets and further using it for organizational decision making. However, finding such exceptions and outliers has not yet received as much attention in the data mining field as some other topics have, such as association rules, classification and clustering. Thus, this paper describes the performance of control chart, linear regression, and Manhattan distance techniques for outlier detection in data mining. Experimental studies show that outlier detection technique using control chart is better than the technique modeled from linear regression because the number of outlier data detected by control chart is smaller than linear regression. Further, experimental studies shows that Manhattan distance technique outperformed compared with the other techniques when the threshold values increased

Soft Set Theoretic Approach for Dimensionality Reduction

A reduct is a subset of attributes that are jointly sufficient and individually necessary for preserving a particular property of a given information system. The existing reduct approaches under soft set theory are still based on Boolean-valued information system. However, in the real applications, the data usually contain non-Boolean valued. In this paper, an alternative approach for attribute reduction in multi-valued information system under soft set theory is presented. Based on the notion of multi-soft sets and AND operation, attribute reduction can be defined. It is shown that the reducts obtained are equivalent with Pawlak’s rough reduction.

Applying variable precision rough set model for clustering student suffering study's anxiety

Computational models of the artificial intelligence such as rough set theory have several applications. Data clustering under rough set theory can be considered as a technique for medical decision making. One possible application is the clustering of student suffering studys anxiety. In this paper, we present the applicability of variable precision rough set model for clustering student suffering studies anxiety. The proposed technique is based on the mean of accuracy of approximation using variable precision of attributes. The datasets are taken from a survey aimed to identify of studies anxiety sources among students at Universiti Malaysia Pahang (UMP). At this stage of the research, we show how variable precision rough set model can be used to groups student in each studys anxiety. The results may potentially contribute to give a recommendation how to design intervention, to conduct a treatment in order to reduce anxiety and further to improve students academic performance.

A Direct Proof of Every Rough Set is a Soft Set

The purpose of this paper is devoted to revealing interconnection between rough sets and soft sets. We use the constructive and descriptive approaches of rough set theory and present a direct proof that Pawlak’s and Iwinski’s rough sets can be considered as soft sets.

Mining significant least association rules using fast SLP-growth algorithm

Development of least association rules mining algorithms are very challenging in data mining. The complexity and excessive in computational cost are always become the main obstacles as compared to mining the frequent rules. Indeed, most of the previous studies still adopting the Apriori-like algorithms which are very time consuming. To address this issue, this paper proposes a scalable trie-based algorithm named SLP-Growth. This algorithm generates the significant patterns using interval support and determines its correlation. Experiments with the real datasets show that the SLP-algorithm can discover highly positive correlated and significant of least association. Indeed, it also outperforms the fast FP-Growth algorithm up to two times, thus verifying its efficiency.

A new linguistic out-sample approach of fuzzy time series for daily forecasting of Malaysian electricity load demand

The fuzzy logical relationships and the midpoints of interval have been used to determine the numerical in-out-samples forecast in the fuzzy time series modeling. However, the absolute percentage error is still yet significantly improved. This can be done where the linguistics time series values should be forecasted in the beginning before the numerical forecasted values obtained. This paper introduces the new approach in determining the linguistic out-sample forecast by using the index numbers of linguistics approach. Moreover, the weights of fuzzy logical relationships are also suggested to compensate the presence of bias in the forecasting. The daily load data from National Electricity Board (TNB) of Malaysia is used as an empirical study and the reliability of the proposed approach is compared with the approach proposed by Yu. The result indicates that the mean absolute percentage error (MAPE) of the proposed approach is smaller than that as proposed by Yu. By using this approach the linguistics time series forecasting and the numerical time series forecasting can be resolved.

Rough Set Approach for Categorical Data Clustering

In this paper, we focus our discussion on the rough set approach for categorical data clustering. We propose MADE (Maximal Attributes Dependency), an alternative technique for categorical data clustering using rough set theory taking into account maximal attributes dependencies. Experimental results on two benchmark UCI datasets show that MADE technique is better with the baseline categorical data clustering techniques with respect to computational complexity and clusters purity.

MAR: Maximum Attribute Relative of soft set for clustering attribute selection

Clustering, which is a set of categorical data into a homogenous class, is a fundamental operation in data mining. One of the techniques of data clustering was performed by introducing a clustering attribute. A number of algorithms have been proposed to address the problem of clustering attribute selection. However, the performance of these algorithms is still an issue due to high computational complexity. This paper proposes a new algorithm called Maximum Attribute Relative (MAR) for clustering attribute selection. It is based on a soft set theory by introducing the concept of the attribute relative in information systems. Based on the experiment on fourteen UCI datasets and a supplier dataset, the proposed algorithm achieved a lower computational time than the three rough set-based algorithms, i.e. TR, MMR, and MDA up to 62%, 64%, and 40% respectively and compared to a soft set-based algorithm, i.e. NSS up to 33%. Furthermore, MAR has a good scalability, i.e. the executing time of the algorithm tends to increase linearly as the number of instances and attributes are increased respectively.

An efficient replicated data access approach for large-scale distributed systems

In data-intensive distributed systems, replication is the most widely used approach to offer high data availability, low bandwidth consumption, increased fault-tolerance and improved scalability of the overall system. Replication-based systems implement replica control protocols that enforce a specified semantics of accessing the data. Also, the performance depends on a host of factors chief of which is the protocol used to maintain consistency among object replica. In this paper, we propose a new low-cost and high data availability protocol for maintaining replicated data on networked distributed computing systems. We show that the proposed approach provides high data availability, low bandwidth consumption, increased fault-tolerance and improved scalability of the overall system as compared to standard replica control protocols.