International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems

In this paper, the weighted quasi-arithmetic means are discussed from the viewpoint of utility functions and downward risks in economics. Representing the weighting functions by probability density functions and the conditional expectations, an index for downward risks in stochastic environments is derived. This paper discusses the relation among the index, the first-order stochastic dominance and the risk premium in economics, and further it investigates the relation between the index and value-at-risks which are known as another estimation for downward risks in finance. Finally, this paper shows a lot of examples of the weighted quasi-arithmetic mean and the aggregated mean ratio for various typical utility functions with various typical utility functions and probability density functions.

In order to analyze the distribution of individual opinions (decision rules) in a group, clustering of decision tables is proposed. An agglomerative hierarchical clustering (AHC) of decision tables has been examined. The result of AHC does not always optimize some criterion. We develop non-hierarchical clustering techniques for decision tables. In order to treat positive and negative evaluations to a common profile, we use a vector of rough membership values to represent individual opinion to a profile. Using rough membership values, we develop a K-means method as well as fuzzy c-means methods for clustering decision tables. We examined the proposed methods in clustering real world decision tables obtained by a questionnaire investigation.

The aim of this paper is to present a new mobile group decision making model to deal with heterogeneous information and changeable decision contexts. This model takes into account that experts have different backgrounds and knowledge levels, allowing to use different preference representations as fuzzy preference relations or linguistic preference relations with multigranular linguistic information. Furthermore, we allow to introduce some changes on the alternatives of the problem at every stage of the decision process. To do that: i) a mobile implementation is proposed to reduce the number of changes and ii) a mechanism to insert/remove alternatives is included in the model. Finally, our new decision model incorporates a feedback mechanism that sends recommendations to the experts in order to quickly obtain a high consensus level.

Fuzzy c-regression models are known to be useful in real applications, but there are two drawbacks: strong dependency on the predefined number of clusters and sensitiveness against outliers or noises. To avoid these drawbacks, we propose sequential fuzzy regression models based on least absolute deviations which we call SFCRMLAD. This algorithm sequentially extracts one cluster at a time using a method of noise-detection, enabling the automatic determination of clusters and having robustness to noises. We compare this method with the ordinary fuzzy c-regression models based on least squares, fuzzy c-regression models based on least absolute deviations, and moreover sequential fuzzy regression models based on least squares. For this purpose we use a two-dimensional illustrative example whereby characteristics of the four methods are made clear. Moreover a simpler and more efficient algorithm of SFCRMLAD can be used for scalar input and output variables, while a general algorithm of SFCRMLAD uses linear programming solutions for multivariable input. By using the above example, we compare efficiency of different algorithms.

Traditional data mining approaches look for patterns in a single table, while multi-relational data mining aims for identifying patterns that involve multiple tables. In recent years, the most common mining techniques have been extended to the multi-relational context, but there are few dedicated to deal with data stored following the multi-dimensional model, in particular the star schema. These schemas are composed of a central huge fact table linking a set of small dimension tables. Joining all the tables before mining may not be a feasible solution due to the usual massive number of records. This work proposes a method for mining frequent patterns on data following a star schema that does not materialize the join between the tables. As it extends the algorithm FP-Growth, it constructs an FP-Tree for each dimension and then combines them through the records in the fact table to form a super FP-Tree. This tree is then mined with FP-growth to find all frequent patterns. The paper presents a case study on bibliographic data, comparing efficiency and scalability of our algorithm against FP-Growth.

Intrusion detection systems play an important role in computer security. To make intrusion detection systems adaptive to changing environments, supervised learning techniques had been applied in intrusion detection. However, supervised learning needs a large amount of training instances to obtain classifiers with high accuracy. Limited to lack of high quality labeled instances, some researchers focused on semi-supervised learning to utilize unlabeled instances enhancing classification. But involving the unlabeled instances into the learning process also introduces vulnerability: attackers can generate fake unlabeled instances to mislead the final classifier so that a few intrusions can not be detected. In this paper we show that the attacker could mislead the semi-supervised intrusion detection classifier by poisoning the unlabeled instances. And we propose a defend method based on active learning to defeat the poisoning attack. Experiments show that the poisoning attack can reduce the accuracy of the semi-supervised learning classifier and the proposed defending method based on active learning can obtain higher accuracy than the original semi-supervised learner under the presented poisoning attack.

In this paper we review the state of the art on graph privacy with special emphasis on applications to online social networks, and we consider some novel aspects which have not been greatly covered in the specialized literature on graph privacy. The following key considerations are covered: (i) choice of different operators to modify the graph; (ii) information loss based on the cost of graph operations in terms of statistical characteristics (degree, clustering coefficient and path length) in the original graph; (iii) computational cost of the operations; (iv) in the case of the aggregation of two nodes, the choice of similar adjacent nodes rather than isomorphic topologies, in order to maintain the overall structure of the graph; (v) a statistically knowledgeable attacker who is able to search for regions of a simply anonymized graph based on statistical characteristics and map those onto a given node and its immediate neighborhood.

The way an author or a group of authors are cited tells more about the real impact of their work than authorship and collaborations. Indeed, the connections within the scientific community can be more accurately elicited from the co-citation graph than from the collaboration graph. We suggest some indices that can be drawn from the co-citation graph in order to capture the relevance of individual authors and the relevance of groups of authors.