Babek G. Guirimov’s research while affiliated with State Oil Company of Azerbaijan Republic and other places

Nowadays, a lot of classification techniques including probabilistic and fuzzy methods exist. The works devoted to dealing with fusion of probabilistic and fuzzy uncertainties of information are scarce. In view of this, partial reliability of information that stems from uncertainty and complexity of real datasets is of interest. Prof. Zadeh introduced a concept of Z-number to describe reliability of information under fuzziness and probabilistic uncertainty. In this work, an approach to Z-number-valued classification of dataset is outlined. The is aim is to describe partial reliability of knowledge expressed by classification. A benchmark data set is used is considered to illustrate the proposed approach.

In the period of internationalization, choice of a country for doing business is a challenging problem. In existing works, this problem is considered for the cases of crisp, statistical, or fuzzy information. However, real international business problems are characterized by fuzziness and partial reliability of information. In this paper we propose a multi-attribute decision making approach to country selection under Z-number-valued information. The approach relies on the methods (which were published by the authors earlier) of consistency-driven Z-number-valued matrix and determination of its eigenvector. A real-world application on country selection with nine real countries is used to illustrate the proposed approach.

The notion of consistency is used to estimate the quality of preference knowledge and its stability for reliable evaluation of decision alternatives. It is well-known that a set of strict consistency conditions are used to keep the rationality of preference intensities between compared elements. These requirements are not achievable in the real situations when decision maker has limited rationality and partially reliable preferences. In this study, we propose an approach to deriving consistency-driven preference degrees for such kind of situations. A preference degree is described by a Z-number to reflect imprecision and partial reliability of preference knowledge. An optimization problem with Z-number valued variables is used to formulate design of consistent preferences. A real-world decision making problem is considered to illustrate application of the proposed method and conduct comparison with an existing technique.

Fuzzy relations were a main tool of fuzzy set theory in decision making, control and other fields. However, partial reliability of decision-relevant information is missed in these approaches. To deal with fuzziness and partial reliability of information, Zadeh introduced the concept of Z-number. The purpose of research presented in this paper is to develop an approach to decision making under Z-number-valued information. We introduce a definition of Z-number-valued relation (Z-relation) and some operations. The reason is to use Z-relations for evaluation of alternatives w.r.t. multiple criteria under imperfect information provided by a decision maker. In view of this, a Z-relation equation is formulated and some results on its solvability are given. These results are a basis of solving decision problems starting from multicriteria evaluation till final ranking of alternatives. The major conclusion is that this approach allows to deal with fusion of fuzzy and probabilistic information at a feasible level of computational complexity. The main limitation of the approach is difficulty of identification of a Z-relation. No expert knowledge (as it requires intensive involvement of experts) or data-driven information (when data quality is low) may exist. At the same time, computational complexity will grow with the high increase of a number of alternatives. A numerical example on decision making for project selection is considered to illustrate applicability of the study.

The paper considers a development of original Differential Evolution optimization algorithm to be used for constrained and multi-objective problems. The paper demonstrates how the considered algorithm and implemented software can be used for solving optimization problems.

A large number of clustering methods exist including deterministic, probabilistic, and fuzzy clustering. All these methods are devoted to handling different types of uncertainty. No studies have been encountered on clustering taking into account a confluence of probabilistic and fuzzy information. In the existing studies, the reliability of extracted knowledge is one of the important issues to be investigated. The concept of Z -number arises as a formal construct that expresses reliability of information under bimodal distribution. In this article, we propose an approach to construction of Z -number-valued clusters of a dataset for evaluation of reliability of extracted data-driven knowledge. Real-world applications are given that confirm the usefulness of the proposed method.

Parkinson's disease is a neurodegenerative movement disorder that causes voice/speech, and behavioral impairments. As a dysfunctional disease, it can be detected by a set of specific symptoms of patients. Such symptoms include both voice/speech and/or physical behavior/movement characteristics. For better detection, both sets of characteristics are used in our research. In this study, as a diagnostic model, we use a system based on multiple-layer (deep) feed-forward neural networks. The networks are trained with Differential Evolution training algorithm using in parallel a pair of data sets (training and validation sets) to avoid overfitting and improve model’s generalization ability (performance on untrained data). The applied DE algorithm has allowed avoiding local minima of error function during the training. A third data set is used for testing trained network performance. According to the obtained results, this method demonstrated better results than other existing approaches.