Publications by Year: 2021

2021
Stanujkić, D., Karabašević, D., Popović, G., Zavadskas, E. K., Saračević, M., Stanimirović, P. S., Katsikis, V. N., et al. (2021). Comparative Analysis of the Simple WISP and Some Prominent MCDM Methods: A Python Approach. Axioms, 10. Publisher's VersionAbstract
This article presents a comparison of the results obtained using the newly proposed Simple Weighted Sum Product method and some prominent multiple criteria decision-making methods. For comparison, several analyses were performed using the Python programming language and its NumPy library. The comparison was also made on a real decision-making problem taken from the literature. The obtained results confirm the high correlation of the results obtained using the Simple Weighted Sum Product method and selected multiple criteria decision-making methods such as TOPSIS, SAW, ARAS, WASPAS, and CoCoSo, which confirms the usability of the Simple Weighted Sum Product method for solving multiple criteria decision-making problems.
Katsikis, V. N., Stanimirovic, P. S., Mourtas, S. D., Xiao, L., Karabasevic, D., & Stanujkic, D. (2021). Zeroing Neural Network with Fuzzy Parameter for Computing Pseudoinverse of Arbitrary Matrix. IEEE Transactions on Fuzzy Systems. Publisher's Version
Khan, A. T., Cao, X., Li, S., Katsikis, V. N., Brajevic, I., & Stanimirovic, P. S. (2021). Fraud detection in publicly traded U.S firms using Beetle Antennae Search: A machine learning approach. Expert Systems with Applications. presented at the 2021. Publisher's VersionAbstract
In this paper, we present a fraud detection framework for publicly traded firms using an optimization approach integrated with a meta-heuristic algorithm known as Beetle Antennae Search (BAS). Existing techniques include human resources, like financial experts and audit teams, to determine the ambiguities or financial frauds in the companies based on financial and non-financial ratios. It is a laborious task, time-consuming, and prone to errors. We designed an optimization problem to minimize the loss function based on a non-linear decision function combined with the maximization of recall (Sensitivity and Specificity). We solved the optimization problem iteratively using the BAS. It is a nature-inspired algorithm and mimics the beetle’s food-searching nature. It includes a single searching particle to find an optimal solution to the optimization problem in n-dimensional space. We used a benchmark dataset collected from SEC’s Accounting and Auditing Enforcement Releases (AAERs) for the simulation. It includes 28 raw financial variables and the data collected between 1991-2008. For the comparison, we evaluated the performance of BAS with the recently proposed approach using RUSBoost. We also compared it with some additional algorithms, i.e., Logit and SVM-FK. The results showed that BAS is comparable with these algorithms and outperformed them in time consumption.
Simos, T. E., Katsikis, V. N., & Mourtas, S. D. (2021). Multi-input bio-inspired weights and structure determination neuronet with applications in European Central Bank publications. Mathematics and Computers in Simulation. presented at the 2021. Publisher's VersionAbstract
This paper introduces a 3-layer feed-forward neuronet model, trained by novel beetle antennae search weights-and-structure-determination (BASWASD) algorithm. On the one hand, the beetle antennae search (BAS) algorithm is a memetic meta-heuristic optimization algorithm capable of solving combinatorial optimization problems. On the other hand, neuronets trained by a weights-and-structure-determination (WASD) algorithm are known to resolve the shortcomings of traditional back-propagation neuronets, including slow speed of training and local minimum. Combining the BAS and WASD algorithms, a novel BASWASD algorithm is created for training neuronets, and a multi-input BASWASD neuronet (MI-BASWASDN) model is introduced. Using a power sigmoid activation function and while managing the model fitting and validation, the BASWASD algorithm finds the optimal weights and structure of the MI-BASWASDN. Four financial datasets, taken from the European Central Bank publications, validate and demonstrate the MI-BASWASDN model’s outstanding learning and predicting performance. Also included is a comparison of the MI-BASWASDN model to three other well-performing neural network models, as well as a MATLAB kit that is publicly available on GitHub to promote and support this research.
Simos, T. E., Katsikis, V. N., & Mourtas, S. D. (2021). A fuzzy WASD neuronet with application in breast cancer prediction. Neural Computing and Applications. presented at the 2021. Publisher's VersionAbstract
Cancer is one of the world’s leading causes of human mortality, and the most prevalent type is breast cancer. However, when diagnosed early, breast cancer may be treated. In this paper, a 5-layer feed-forward neuronet model, trained by a novel fuzzy WASD (weights-and-structure-determination) algorithm, called FUZWASD, is introduced and employed to predict whether the breast cancer is benign or malignant. In general, WASD-trained neuronets are known to overcome the limitations of traditional back-propagation neuronets, including slow training speed and local minimum; however, multi-input WASD-trained neuronets with no dimension explosion weakness are few. In this work, a novel FUZWASD algorithm for training neuronets is modeled by embedding a fuzzy logic controller (FLC) in a WASD algorithm, and a multi-input FUZWASD neuronet (MI-FUZWASDN) model for classification problems with no dimension explosion weakness is proposed. The FUZWASD algorithm uses a FLC to map the input data into a specific interval that enhances the accuracy of the weights-direct-determination (WDD) method. In this way, the FUZWASD algorithm detects the optimal weights and structure of the MI-FUZWASDN using a power softplus activation function and while handling the model fitting and validation. Applications on two diagnostic breast cancer datasets validate and demonstrate the MI-FUZWASDN model’s exceptional learning and predicting performance. In addition, for the intrigued user, we have created a MATLAB kit, which is freely accessible via GitHub, to promote and support the results of this work.
Mourtas, S. D., & Katsikis, V. N. (2021). V-Shaped BAS: Applications on Large Portfolios Selection Problem. Computational Economics. presented at the 2021. Publisher's VersionAbstract
The beetle antennae search (BAS) algorithm is a memetic meta-heuristic optimization algorithm capable of solving combinatorial optimization problems. In this paper, the binary version of BAS (BBAS) is modified by adding a V-shaped transfer function. In this way, we introduce the V-shaped transfer function-based binary BAS (VSBAS) algorithm, which is a more effective and efficient version of BBAS in the case of large input data. Applications using real-world data sets on a binary Markowitz-based portfolio selection (BMPS) problem validate the excellent performance of VSBAS on large input data and demonstrate that it is a marvelous alternative against other ordinary memetic meta-heuristic optimization algorithms. Note that, because the meta-heuristic algorithms compared in this paper are directly applicable only to unconstrained optimization, the penalty function method was used to keep their solutions in the feasible district. In order to support and promote the findings of this work, we have constructed a complete MATLAB package for the interested user, which is freely available through GitHub.
Simos, T. E., Mourtas, S. D., & Katsikis, V. N. (2021). Time-varying Black–Litterman portfolio optimization using a bio-inspired approach and neuronets. Applied Soft Computing, 112, 107767. WebsiteAbstract
The Black–Litterman model is a very important analytical tool for active portfolio management because it allows investment analysts to incorporate investor’s views into market equilibrium returns. In this paper, we define and study the time-varying Black–Litterman portfolio optimization under nonlinear constraints (TV-BLPONC) problem as a nonlinear programming (NLP) problem. More precisely, the nonlinear constraints refer to transaction costs and cardinality constraints. Furthermore, a speedy weights-and-structure-determination (WASD) algorithm for the power-activation feed-forward neuronet (PFN) is presented to solve time-series modeling and forecasting problems. Inhere, the investor’s views in the TV-BLPONC problem are considered as a forecasting problem and, thus, they are produced by the WASD-based PFN. In addition, using the beetle antennae search (BAS) algorithm a computational method is introduced to solve the TV-BLPONC problem. For all we know, this is an innovative approach that integrates modern neural network and meta-heuristic optimization methods to provide a solution to the TV-BLPONC problem in large portfolios. Our approach is tested on portfolios of up to 90 stocks with real-world data, and the results show that it is more than 30 times faster than other methods. Our technique’s speed and precision are verified in this way, showing that it is an outstanding alternative to ordinary methods. In order to support and promote the findings of this work, we have constructed two complete MATLAB packages for the interested user, which are freely available through GitHub.
Mosić, D., Stanimirović, P. S., & Katsikis, V. N. (2021). Weighted composite outer inverses. Applied Mathematics and Computation, 411, 126493. WebsiteAbstract
In order to extend and unify the definitions of W-weighted DMP, W-weighted MPD, W-weighted CMP and composite outer inverses, we present the weighted composite outer inverses. Precisely, the notions of MNOMP, MPMNO and MPMNOMP inverses are introduced as appropriate expressions involving the (M,N)-weighted (B,C)-inverse and Moore–Penrose inverse. Basic properties and a number of characterizations for the MNOMP, MPMNO or MPMNOMP inverse are discovered. Various representations and characterizations of weighted composite outer inverses are studied. General solutions for certain systems of linear equations are given in terms of weighted composite outer inverses. Numerical examples are presented on randomly generated matrices of various orders.
Stanujkić, D., Karabašević, D., Popović, G., Stanimirović, P. S., Saračević, M., Smarandache, F., Katsikis, V. N., et al. (2021). A New Grey Approach for Using SWARA and PIPRECIA Methods in a Group Decision-Making Environment. Mathematics, 9. Publisher's VersionAbstract
The environment in which the decision-making process takes place is often characterized by uncertainty and vagueness and, because of that, sometimes it is very hard to express the criteria weights with crisp numbers. Therefore, the application of the Grey System Theory, i.e., grey numbers, in this case, is very convenient when it comes to determination of the criteria weights with partially known information. Besides, the criteria weights have a significant role in the multiple criteria decision-making process. Many ordinary multiple criteria decision-making methods are adapted for using grey numbers, and this is the case in this article as well. A new grey extension of the certain multiple criteria decision-making methods for the determination of the criteria weights is proposed. Therefore, the article aims to propose a new extension of the Step-wise Weight Assessment Ratio Analysis (SWARA) and PIvot Pairwise Relative Criteria Importance Assessment (PIPRECIA) methods adapted for group decision-making. In the proposed approach, attitudes of decision-makers are transformed into grey group attitudes, which allows taking advantage of the benefit that grey numbers provide over crisp numbers. The main advantage of the proposed approach in relation to the use of crisp numbers is the ability to conduct different analyses, i.e., considering different scenarios, such as pessimistic, optimistic, and so on. By varying the value of the whitening coefficient, different weights of the criteria can be obtained, and it should be emphasized that this approach gives the same weights as in the case of crisp numbers when the whitening coefficient has a value of 0.5. In addition, in this approach, the grey number was formed based on the median value of collected responses because it better maintains the deviation from the normal distribution of the collected responses. The application of the proposed approach was considered through two numerical illustrations, based on which appropriate conclusions were drawn.
Katsikis, V. N., Mourtas, S. D., Stanimirović, P. S., & Zhang, Y. (2021). Continuous-Time Varying Complex QR Decomposition via Zeroing Neural Dynamics. Neural Processing Letters. presented at the 2021. Publisher's VersionAbstract
QR decomposition (QRD) is of fundamental importance for matrix factorization in both real and complex cases. In this paper, by using zeroing neural dynamics method, a continuous-time model is proposed for solving the time-varying problem of QRD in real-time. The proposed dynamics use time derivative information from a known real or complex matrix. Furthermore, its theoretical analysis is provided to substantiate the convergence and effectiveness of solving the time-varying QRD problem. In addition, numerical experiments in four different-dimensional time-varying matrices show that the proposed model is effective for solving the time-varying QRD problem both in the case of a real or a complex matrix as input.
Stanujkić, D., Karabašević, D., Popović, G., Stanimirović, P. S., Smarandache, F., Saračević, M., Katsikis, V. N., et al. (2021). An Innovative Grey Approach for Group Multi-Criteria Decision Analysis Based on the Median of Ratings by Using Python. Axioms, 10. Publisher's VersionAbstract
Some decision-making problems, i.e., multi-criteria decision analysis (MCDA) problems, require taking into account the attitudes of a large number of decision-makers and/or respondents. Therefore, an approach to the transformation of crisp ratings, collected from respondents, in grey interval numbers form based on the median of collected scores, i.e., ratings, is considered in this article. In this way, the simplicity of collecting respondents’ attitudes using crisp values, i.e., by applying some form of Likert scale, is combined with the advantages that can be achieved by using grey interval numbers. In this way, a grey extension of MCDA methods is obtained. The application of the proposed approach was considered in the example of evaluating the websites of tourism organizations by using several MCDA methods. Additionally, an analysis of the application of the proposed approach in the case of a large number of respondents, done in Python, is presented. The advantages of the proposed method, as well as its possible limitations, are summarized.
Katsikis, V. N., & Mourtas, S. D. (2021). Binary Beetle Antennae Search Algorithm for Tangency Portfolio Diversification. Journal of Modeling and Optimization, 13(1). Publisher's VersionAbstract
The tangency portfolio, also known as the market portfolio, is the most efficient portfolio and arises from the intercept point of the Capital Market Line (CML) and the efficient frontier. In this paper, a binary optimal tangency portfolio under cardinality constraint (BOTPCC) problem is defined and studied as a nonlinear programming (NLP) problem. Because such NLP problems are widely approached by heuristic, a binary beetle antennae search algorithm is employed to provide a solution to the BTPSCC problem. Our method proved to be a magnificent substitute to other evolutionary algorithms in real-world datasets, based on numerical applications and computer simulations.
Katsikis, V. N., & Mourtas, S. D. (2021). Portfolio Insurance and Intelligent Algorithms. In S. Patnaik, Tajeddini, K., & Jain, V. (Eds.), Computational Management: Applications of Computational Intelligence in Business Management (pp. 305 - 323). presented at the 2021, Cham: Springer International Publishing. Publisher's VersionAbstract
Minimizing portfolio insurance (PI) costs is an investment strategy of great importance. In this chapter, by converting the classical minimum-cost PI (MCPI) problem to a multi-period MCPI (MPMCPI) problem, we define and investigate the MPMCPI under transaction costs (MPMCPITC) problem as a nonlinear programming (NLP) problem. The problem of MCPI gets more genuine in this way. Given the fact that such NLP problems are widely handled by intelligent algorithms, we are introducing a well-tuned approach that can solve the challenging MPMCPITC problem. In our portfolios’ applications, we use real-world data and, along with some of the best memetic meta-heuristic and commercial methods, we provide a solution to the MPMCPITC problem, and we compare their solutions to each other.
Stanimirović, P. S., Katsikis, V. N., Jin, L., & Mosić, D. (2021). Properties and computation of continuous-time solutions to linear systems. Applied Mathematics and Computation, 405. presented at the 2021. Publisher's VersionAbstract
We investigate solutions to the system of linear equations (SoLE) in both the time-varying and time-invariant cases, using both gradient neural network (GNN) and Zhang neural network (ZNN) designs. Two major limitations should be overcome. The first limitation is the inapplicability of GNN models in time-varying environment, while the second constraint is the possibility of using the ZNN design only under the presence of invertible coefficient matrix. In this paper, by overcoming the possible limitations, we suggest, in all possible cases, a suitable solution for a consistent or inconsistent linear system. Convergence properties are investigated as well as exact solutions.
Katsikis, V. N., Mourtas, S. D., Stanimirović, P. S., Li, S., & Cao, X. (2021). Time-Varying Mean-Variance Portfolio Selection under Transaction Costs and Cardinality Constraint Problem via Beetle Antennae Search Algorithm (BAS). Operations Research Forum, 2(2), 18. presented at the 2021. Publisher's VersionAbstract
The Markowitz mean-variance portfolio selection is widely acclaimed as a very important investment strategy. A popular option to solve the static mean-variance portfolio selection (MVPS) problem is based on the use of quadratic programming (QP) methods. On the other hand, the static portfolio selection under transaction costs (PSTC) problem is usually approached with nonlinear programming (NLP) methods. In this article, we define and study the time-varying mean-variance portfolio selection under transaction costs and cardinality constraint (TV-MVPSTC-CC) problem as a time-varying nonlinear programming (TVNLP) problem. The TV-MVPSTC-CC also comprises the properties of a moving average. These properties make the TV-MVPSTC-CC an even greater analysis tool suitable to evaluate investments and identify trading opportunities across a continuous-time period. Using the Beetle Antennae Search (BAS) algorithm, we also provide an online solution to the static NLP problem. To the best of our knowledge, this is an innovative approach that incorporates modern meta-heuristic optimization techniques to provide an online, thus more realistic, solution to the TV-MVPSTC-CC problem. In this way, we present an online solution to a time-varying financial problem while eliminating the restrictions of static methods. Our approach is also verified by numerical experiments and computer simulations as an excellent alternative to traditional approaches.
Li, Z., Zhang, Y., Ming, L., Guo, J., & Katsikis, V. N. (2021). Real-Domain QR Decomposition Models Employing Zeroing Neural Network and Time-Discretization Formulas for Time-Varying Matrices. Neurocomputing. presented at the 2021. Publisher's VersionAbstract
This study investigated the problem of QR decomposition for time-varying matrices. We transform the original QR decomposition problem into an equation system using its constraints. Then, we propose a continuous-time QR decomposition (CTQRD) model by applying zeroing neural network method, equivalent transformations, Kronecker product, and vectorization techniques. Subsequently, a high-precision ten-instant Zhang et al discretization (ZeaD) formula is proposed. A ten-instant discrete-time QR decomposition model is also proposed by using the ten-instant ZeaD formula to discretize the CTQRD model. Moreover, three discrete-time QR decomposition models are proposed by applying three other ZeaD formulas, and three examples of QR decomposition are presented. The experimental results confirm the effectiveness and correctness of the proposed models for the QR decomposition of time-varying matrices.
Katsikis, V. N., Mourtas, S. D., Stanimirović, P. S., & Zhang, Y. (2021). Solving Complex-Valued Time-Varying Linear Matrix Equations via QR Decomposition With Applications to Robotic Motion Tracking and on Angle-of-Arrival Localization. IEEE Transactions on Neural Networks and Learning Systems, 1 - 10.