International Journal of Information Technology & Decision Making

An unsupervised feature selection method is proposed for analysis of datasets of high dimensionality. The least square error (LSE) of approximating the complete dataset via a reduced feature subset is proposed as the quality measure for feature selection. Guided by the minimization of the LSE, a kernel least squares forward selection algorithm (KLS-FS) is developed that is capable of both linear and non-linear feature selection. An incremental LSE computation is designed to accelerate the selection process and, therefore, enhances the scalability of KLS-FS to high-dimensional datasets. The superiority of the proposed feature selection algorithm, in terms of keeping principal data structures, learning performances in classification and clustering applications, and robustness, is demonstrated using various real-life datasets of different sizes and dimensions.

During the last decade, information system (IS) outsourcing has emerged as a major issue for organizations. As outsourcing decisions are often based on multicriteria approaches and group decisions, this paper proposes a structured methodology based on Fuzzy group decision making approach to evaluate and select the appropriate information system project (ISP) in an actual case. To achieve our purpose, we argue that seven criteria consisting of risk, management, economics, technology, resource, quality, and strategy and five ISPs should be considered for outsourcing decisions. Fuzzy analytic hierarchy process (fuzzy AHP) to find the priority and ranking of each ISP is considered. Using fuzzy theory for selecting the proper IS project can reduce ambiguities and uncertainties that are inherent in the selection procedure. A sensitivity analysis is performed to check the steadiness of the priority ranking and help decision makers to understand different scenarios that show alternative potential developments or different viewpoints concerning the relative importance of the criteria. Our proposed methodology can be applied by a cross-functional team of practitioners and IT managers to select the proper IS project. Also, the proposed approach can effectively consider complex and qualitative decision variables involved in the IS outsourcing decision-making problem. Finally, conclusion and potential issues for future research are presented. The main findings of this research have showed that the combination of AHP method and fuzzy concept to rank and prioritize of IS projects to outsource is a useful and practical tool to make traceable and reliable decision. According to results, a priority of overall score for the five candidates of IS projects reflects that project of "development of the supplier relationship management information system" performs the best, and project of "facilities management" performs the worst according to the experts' judgment in case study.

This article focuses on the integration of multicriteria decision analysis (MCDA) and geographical information systems (GIS) and introduces a tool, GIS–MCDA, written in visual basic in ArcGIS for GIS-based MCDA. The GIS–MCDA deals with raster-based data sets and includes standardization, weighting and decision analysis methods, and sensitivity analysis. Simple additive weighting, weighted product method, technique for order preference by similarity to ideal solution, compromise programming, analytic hierarchy process, and ordered weighted average for decision analysis; ranking, rating, and pairwise comparison for weighting and linear scale transformation for standardization can be applied by using this tool. The maximum score and score range procedures can be used for linear scale transformation. In this article also an application of the GIS–MCDA to determine the flood vulnerability of the South Marmara Basin in Turkey is examined. To check the validity and reliability of the results, the flood vulnerability layer is compared with flood-affected areas.

Consistency plays an important part in decision making with interval reciprocal relation also called interval fuzzy preference relation. In this paper, we give the definition of perfect multiplicative consistent interval reciprocal relation and discuss its properties. A method is proposed to construct a set of multiplicative consistent reciprocal relations from a perfect consistent interval reciprocal relation. Then we develop methods to get the perfect multiplicative consistent interval reciprocal relation from the inconsistent one and estimate the missing values from an incomplete interval reciprocal relation. Finally, a group decision-making method is proposed based on perfect multiplicative consistent interval reciprocal relations. Some examples are also given to compare the proposed methods with the existing ones.

The accuracy of the results obtained by using multicriteria evaluation methods largely depends on the determination of the criteria weights. The accuracy of expert evaluation decreases with the increase of the number of criteria. The application of the analytic hierarchy process, based on pairwise comparison of the criteria, or similar methods, may help to solve this problem. However, if the number of the pairs of criteria is large, the same problems, associated with the accuracy of evaluation, arise. In the present paper, a new method of determining the criteria weights, FARE (Factor Relationship), based on the relationships between all the criteria describing the phenomenon considered, is offered. It means that, at the first stage, a minimal amount of the initial data about the relationships between a part of the set of criteria, as well as their strength and direction, is elicited from experts. Then, based on the conditions of functioning and the specific features of the complete set of criteria, the relations between other criteria of the set and their direction are determined analytically in compliance with those established at the first stage. When the total impact of each particular criterion on other criteria of the set or its total dependence on other criteria of a particular set is known, the criteria weights can be determined.

Sentiment classification seeks to identify general attitude of a piece of text of comments or reviews on certain subject, be it positive or negative. Most existing researches on sentiment classification employ supervised learning approaches that rely on annotated data. However, sentiment is expressed differently on different subjects in different domains, and having annotated corpora for every domain of interest is not always practical. This paper proposes an unsupervised learning approach for classifying text of online reviews as recommended or not recommended. The proposed method is based on search engine snippet, summary information on the result page of a search engine. A basic assumption is that terms with similar orientation tend to co-occur. The co-occurrence is measured by utilizing snippets returned from search engines, with a query consisting of the text and a seed positive or negative word. With the information of snippets, the proposed method may estimate the association of candidate terms more accurately. This allows us to reliably predict the sentiment orientation of customer reviews. Texts of customer reviews are then classified as recommended or not recommended if the average sentiment orientations of its phrases are positive or negative. The research data set of this study consists of 600 Chinese online reviews about travel destinations retrieved from Ctrip.com. Our approach achieves an accuracy of 76.5%. Factors that influence the accuracy of the sentiment classification of Chinese online reviews were discussed.

In this paper, we study group decision-making problems based on intuitionistic preference relations. By measuring the uncertain information of intuitionistic preference relations and the average similarity degree of one individual intuitionistic preference relation to the others, we propose a new approach to assess the relative importance weights of experts. The approach takes both the objective and subjective information of experts into consideration. We then integrate the weights of experts into the individual intuitionistic preference relations and develop a relative similarity method to derive the priorities of alternatives. The comparison analysis with other methods by two numerical examples illustrates the practicality and effectiveness of the proposed methods.

The theory of interval valued fuzzy sets is very valuable for modeling impressions of decision makers. In addition, it gives ability to quantify the ambiguous nature of subjective judgments in an easy way. In this paper, by extending the technique for order preference by similarity to ideal solution (TOPSIS), it is proposed a useful method based on generalized interval valued trapezoidal fuzzy numbers (GITrFNs) for solving multiple criteria decision analysis (MCDA) problems. In view of complexity in handling sophisticated data of GITrFNs, this paper employs the concept of signed distances to establish a simple and effective MCDA method based on the main structure of TOPSIS. An algorithm based on TOPSIS method is established to determine the priority order of given alternatives by using properties of signed distances. Finally, the feasibility of the proposed method is illustrated by a practical example of supplier selection.

Multiple criteria linear programming and multiple criteria quadratic programming classification models have been applied in some field in financial risk analysis and credit risk control such as credit cardholders' behavior analysis. In this paper, a fuzzy linear programming classification method with soft constraints and criteria was proposed based on the previous findings from other researchers. In this method, the satisfied result can be obtained through selecting constraint and criteria boundary variable d_i*, respectively. A general framework of this method is also constructed. Two real-life datasets, one from a major USA bank and the other from a database of KDD 99, are used to test the accurate rate of the proposed method. And the result shows the feasibility of this method.

This paper treats the infinite horizon discounted cost control problem for partially observable Markov decision processes. Sondik studied the class of finitely transient policies and showed that their value functions over an infinite time horizon are piecewise linear (p.w.l) and can be computed exactly by solving a system of linear equations. However, the condition for finite transience is stronger than is needed to ensure p.w.l. value functions. In this paper, we introduce alternatively the class of periodic policies whose value functions turn out to be also p.w.l. Moreover, we examine a more general condition than finite transience and periodicity that ensures p.w.l. value functions. We implement these ideas in a replacement problem under Markovian deterioration, investigate for periodic policies and give numerical examples.