This paper proposes a new fuzzy k-nearest neighbor (FKNN) method, called the ordered weighted averaging (OWA) with regular increasing monotone quantifier-based fuzzy k-nearest neighbor (OWARIM-FKNN) classifier. The proposed method aims at enhancing the classification performance of the KNN rule-base variants, especially the local mean-based approaches, while dealing with outlier and data uncertainty issues. In the proposed method, the OWA operator is used to generalize the multi-local mean vectors from each class. The resulting k multi-local OWA vectors are then used to create the class representative pseudo nearest neighbors. Lastly, the new sample is classified into the class with the highest membership degree measured using the weighted distance between the new sample and the pseudo nearest neighbor. The classification performance of the proposed method was examined using one artificial and twenty-seven real-world data sets compared with the results obtained from eight related KNN variants. Experimental results showed that the proposed OWARIM-FKNN classifier achieves the highest average accuracy of 87.59% with an average confidence interval of ± 0.64, outperforming all baseline methods. Using the Friedman and Nemenyi tests, the analysis further confirms that the proposed method shows statistically significant performance improvements. • A new variant of FKNN method using an OWA operator and a RIM quantifier is proposed. • Nearest neighbors-based multi-local OWA vectors are introduced in the FKNN method. • The new classifier is more robust to outlier effects than the state-of-the-art methods used. • Results highlight the effectiveness of the proposed approach for classification problems.

This study proposes an application of cognitive maps in the representation of cognitive structures of the experts and assessment of their development/modification as a result of a (computer or expert system-assisted) learning process. It strives to identify information needed for the guidance of the process of creation and management of expert knowledge by formal modeling tools. Changes in experts’ cognitive structures are assumed to stem from individual and collaborative (group-level) learning. The novel approach to assessing the outcomes of learning reflected as changes in the cognitive structures of experts or groups of experts, modeled by cognitive maps, does not assume any correct or desired outcome of the learning process to be known in advance. Instead, it identifies and analyses the changes in (or robustness of) the constituents of the cognitive maps from different points of view and allows for quantifying and visualizing the actual effect of the learning. The proposed methodology can identify changes in cognitive diversity, causal structures in terms of causal relations and concepts, and the perceived importance of strategic issues over the learning period. It can also detect which cause–effect relationships have appeared/disappeared considering the pre-/post-mapping design. Thus it provides an exploratory account on the changes in the cognitive structures of the expert(s) as a result of learning. The applicability of the proposed methods is illustrated in the assessment of the learning outcomes of a group of 71 graduate students who participated in an eight-week business simulation task. The results of the empirical analysis confirm the viability of the proposed methodology and indicate that the students’ understanding of the utilized concepts and associated relationships in the decision-making process improved throughout the learning activity, ultimately showing that the course learning has considerably improved students’ perception and knowledge. Based on the results, it can be concluded that the proposed approach has the potential to be effective in assessing learning outcomes in a teaching-learning activities.

This paper describes a set of proposed additions to the recently introduced possibilistic fuzzy pay-off method for real option valuation that enhances the practical usability of the method. The additions are focused on the practical usability of the method and concentrate on emphasizing the consideration of the downside-risk found in projects. Technically the additions are based on using a novel interpretation of the possibilistic mean as a proxy respectively for the weight of the downside and the upside of a possibility distribution in the context of project profitability. This interpretation of the possibilistic mean is a new theoretical contribution. The proposed new method-variant, called the “practical possibilistic fuzzy pay-off method for real option valuation”, can effectively distinguish between projects with an identical upside and non-identical downsides and allows for a more finance-theoretically comprehensive consideration of situations, where the circumstances surrounding the downside risk of a project change. Changes in the downside are reflected in the real option value. The proposed changes constitute the first variant of the possibilistic fuzzy pay-off method for real option valuation and they remarkably increase the practical usability of the method.

This paper investigates the formulation of the dissemblance index as a basis for the calculation of distances of fuzzy numbers and explores its potential linkages with standard and possibilistic moments of fuzzy numbers. Applying the LSC transformation introduced recently by Luukka, Stoklasa and Collan we transform the general formulation of the dissemblance index into its “probabilistic” analogy and show that the result can be interpreted as a difference of COGs of the respective fuzzy numbers (potentially with hedges applied to them). We also show that the difference of possibilistic means is a special case of the general dissemblance index, when w=1. We also propose a generalized version of the possibilistic mean of a fuzzy number and prove its properties. We discuss the implications of this relationship on the practical use of the generalized dissemblance index and investigate its performance in the task of ranking of fuzzy numbers.

This paper introduces a new type of mean applicable in various areas of science and practice: the α-weighted averaging operator (AWA). AWA has all the properties required from a linear averaging operator and some additional ones. We discuss the applications of AWA in data aggregation in various areas including uncertainty modeling (summarization, defuzzification), multiple-criteria and multi-expert decision-making and evaluation. We prove that when applied to fuzzy numbers, the α-weighted average converges to the possibilistic mean of a fuzzy number with the increasing number of elements in its support. As such the α-weighted average is a more general aggregation operator than the original possibilistic mean. When fuzzy subsets of the real line represent the information to be aggregated, AWA provides new means for their defuzzification compatible with the possibilistic moments, but applicable to discrete and subnormal fuzzy sets. We also introduce a generalized formulation of the α-weighted averaging operator (GAWA) that can be applied in multiple-criteria and multi-expert evaluation and decision-making problems. We suggest the use of GAWA in operations research theory and applications in the context of data aggregation, multiple-criteria and group evaluation and decision-making.

In this paper we propose three novel feature ranking methods for supervised feature selection in the context of classification which are based on possibility theory. All three methods – nonspecificity, strife and total uncertainty – are tested on eight artificial data sets and ten medical real-world data sets and benchmarked against ReliefF, the Fisher score, the fuzzy entropy and similarity (FES), the Fuzzy similarity and entropy (FSAE) filter, symmetrical uncertainty as well as using no feature selection. The feature ranking methods were applied following two approaches: (1) using a fixed threshold for the number of highest-ranking features selected and (2) using a hybrid feature selection approach with a classifier (k-nearest neighbor classifier, decision tree, similarity classifier, SVM) to select the optimal number of features to select. The results indicate that strife and the Fisher score are the two feature ranking methods that for both approaches are on average ranked the highest in terms of the test set accuracy on the real-world data sets. Besides that, for the hybrid approach, strife uses most of the time a considerably smaller number of features than nonspecificity and total uncertainty. In terms of stability, which was measured with the adjusted stability measure (ASM), the Fisher score and strife were among the most stable feature ranking methods in this study. Additionally, strife’s feature subsets were diverse compared to those of the remaining feature selection methods, making it a good candidate to be included in a feature selection ensemble.

In this literature review, we investigate machine learning techniques that are applied for stock market prediction. A focus area in this literature review is the stock markets investigated in the literature as well as the types of variables used as input in the machine learning techniques used for predicting these markets. We examined 138 journal articles published between 2000 and 2019. The main contributions of this review are: (1) an extensive examination of the data, in particular, the markets and stock indices covered in the predictions, as well as the 2173 unique variables used for stock market predictions, including technical indicators, macro-economic variables, and fundamental indicators, and (2) an in-depth review of the machine learning techniques and their variants deployed for the predictions. In addition, we provide a bibliometric analysis of these journal articles, highlighting the most influential works and articles.

New Technique for Order Preference by Similarity to Ideal Solution, (TOPSIS), variant is presented, where n-ary norm operators are used in creating ideal solutions. We show that we are gaining different ranking results with this new proposal. Also we address reasons for why this variant is giving different ranking orders compared to original TOPSIS and propose a way to try to select suitable n-ary norm operations and how to try to select a suitable parameter in case of parametrized norm operators. New method is examined with the patent selection problem.

In this paper we introduce a generalization of the similarity based TOPSIS by using the Bonferroni mean. By generalizing similarity computation between an alternative and the ideal solution we also create an effect, where the ranking of alternatives can change depending on parameter value selection. For this purpose we apply the histogram ranking method to take into consideration the variability of rankings for a more robust holistic ranking of the alternatives. The proposed method is applied to a patent portfolio selection problem.

Research and development (R&D) project ranking as investments is a well-known problem that is made difficult by incomplete and imprecise information about future project profitability. This paper shows how profitability results of R&D project evaluation with the fuzzy pay-off method can be ranked with four new variants of fuzzy TOPSIS each using a different fuzzy similarity measure. An overall project ranking that incorporates the four new variants' rankings with three different ideal solutions totaling 12 subrankings is presented. The implementation of the created methods is illustrated with a numerical example.

A novel procedure for human resources selection is proposed. Fuzzy scorecards are used to collect information from multiple experts representing different knowledge domains. The results are aggregated by using Fuzzy Heavy Ordered Weighted Averaging (FHOWA) total type aggregation. The results are ranked with a new method, based on using fuzzy similarity to an ideal solution. A numerical example is used for illustration. The procedure preserves more information from evaluation to ranking than previous models.

In this article we propose a differential evolution based nearest prototype classifier with extension to selecting the applied distance measure from a pool of alternative measures optimally for the particular data set at hand. The proposed method extends the earlier differential evolution based nearest prototype classifier by extending the optimization process to cover also the selection of distance measure instead of optimizing only the parameters related with a preselected and fixed distance measure. Now the optimization process is seeking also for the best distance measure providing the highest classification accuracy over the selected data set. It has been clear for some time that in classification, the usual euclidean distance measure is sometimes not the best possible choice. Still usually not much has been done for it, and in many cases where some consideration to this problem is given, there has only been testing with a couple of alternative distance measures to find which one provides the highest classification accuracy over the current data set. In this paper we attempt to take one step further by not only enumerating a couple of alternative distance measures, but applying a systematic optimization process to select the best distance measure from a pool of multiple alternative distance measures. In parallel, within the same optimization process, the optimal parameter values related to each alternative distance measures are determined as well as the optimal class prototype vectors for the given data. The empirical results represented are indicating that with several data sets the optimal distance measure is some other measure than the most commonly applied euclidean distance. The results are also suggesting that from the classification accuracy point of view the proposed global optimization approach has high potential in solving classification problems of the studied type. Perhaps the most generally applicable conclusion from our results is, that emphasizing of selection of distance measure is more important to classification accuracy that it has been commonly believed so far.

In classification problems feature selection has an important role for several reasons. It can reduce computational cost by simplifying the model. Also when the model is taken for practical use fewer inputs are needed which means in practice, that fewer measurements from new samples are needed. Removing insignificant features from the data set makes the model more transparent and more comprehensible. In this way the model can be used to provide better explanation to the medical diagnosis, which is an important requirement in medical applications. Feature selection process can also reduce noise, this way enhancing the classification accuracy. In this article feature selection method based similarity measure using Yu's similarity with fuzzy entropy measures is introduced and it is tested together with the similarity classifier. Model was tested with dermatology data set. When comparing the results to previous works the results compare quite well. Mean classification accuracy with dermatology data set was 98.83% and it was achieved using 33 features instead of 34 original features. Results can be considered quite good.

A general fuzzy linear system is investigated using fuzzy numbers and Gauss-Seidel iteration formula. We have used our fuzzy linear system to solve Leontief input-output model with fuzzy entries. When solving Leontief input-output model we are usually making the assumption that we know entirely the consumption matrix from industrial entries and we are certain about the final demand. These assumptions however depend heavily on estimates and information received from the industry and hence in these estimates, uncertainty plays a crucial role. To address this type of uncertainty fuzzy methods are needed to model this and in this article we are giving a procedure to solve this problem. Numerical example is given to illustrate the procedure.

In this paper classification results using similarity classifier and Yu's norms are applied to medical data sets. Similarity measure is constructed using measure based on Yu's t-norm and t-conorm. Constructed measure has been succesfully used in classification. Also results are compared to DIMLP, MLP and CN2 classifiers which are giving quite good results. Result presented in this paper are quite promising. The effects of PCA and entropy minimization as preprocessing methods to classification results has been studied. Results show that entropy minimization method worked well as preprocessor with these data sets and similarity classifier.

In this paper a study of similarity based classifier with generalized mean and different preprocessors suites for medical data sets. Also results are compared to C4.5 decision trees, multi-layer perceptrons and DIMLP which are giving quite good results. Result presented in this paper are promising. Also wider testing of classifier based on generalized mean base fuzzy similarity is carried out. The effects of PCA and entropy minimization as preprocessing methods to classification results with similarity based classifier is conducted.

We have investigated effects of different dimension reduction methods to fuzzy similarity based classifier. We are going to show that dimension reduction plays an important role especially with geometric and harmonic similarity based classification. We are also going to show that the arithmetic case does not depend that much on dimension reduction, but very similar results can be achieved with full dimension instead of optimal dimension.

In this paper, we have done new similarity measures from a continuous t-norm by implementing it in different mean measures. For the implementation, we use a Minkowsky metric based on Lukasiewicz algebra. We test these new similarities in both the generalised and normal form of Lukasiewicz algebra with weight optimisation. The mean measures examined here are arithmetic, geometric and harmonic means. We show that the magnitude order of the similarities are S/sub H//sup N//spl ges/S/sub G//sup N//spl ges/S/sub A//sup N/. Secondly, we show that the use of different means is highly recommendable in some cases.

The first aim of this paper is to extend the fuzzy similarity relation defined in the generalized Lukasiewicz structure to utilize common and cumulative Minkowsky metrics and introduce a new classifier based on the cumulative similarity measure which uses the Minkowsky metrics in the generalized Lukasiewicz structure. The second aim of this paper is to study properties of this classifier when applied to three different datasets.

The aim of this paper is to introduce improvements made to a classifier based on maximal fuzzy similarity. Improvements are based on the use of generalized Lukasiewicz-structure and weight optimization. The main benefits of the classifier are its computational efficiency and its strong mathematical background. It is based on many-valued logic and it provides semantic information about classification results. We show that if one chooses the power value in a right manner in the generalized Lukasiewicz-structure and the optimal weights for different feature, one can see significant enhancements in classification results.