Dmytro Peleshko

GeoGuard

The basis of any business is customer databases, which provide information on customer relations with the company. For example, in the field of banking services, the database stores information about the client, account number, data on financial transactions in online trading, purchased goods, their quantity, time of purchase, etc. A fraud detectio...

The Hybrid Deep Convolutional Neural Network with Multimodal Fusion (HDCNNMF) topology for the multimodal recognition of the speech, the face, the lips, and human gestures behavior is proposed. Conducted researches relate to improving the understanding of complex dynamic scenes. The basic unit of the proposed hybrid system is deep neural network to...

Time series arise in different fields of the economy and forecasting of them is important part of decision making. However, their intrinsic complexity and nonlinear behavior makes prediction in that field a challenging task. Hybrid artificial intelligence models are among the most powerful tools in handling such complex dynamics. This paper introdu...

Nowadays the real-time object tracking in video streams is an important problem in smart city applications. One of the problem of smart city monitoring such as the flame detection task in real time is considered. The flame as the object of interest is one of the most complex objects tracking. This is due to the lack of time and space invariance in...

Since the airspace is subordinated to the territories of different states, different authorities, such as Eurocontrol in Europe and FAA in America, etc., an acute need for a product that would provide easy, fast, high-quality flight planning, their proper dissemination, formation of necessary documents, etc. arises. Purpose is to cluster all necess...

AHP study methods of solving problems to use AHP comparing alternatives (TOPSIS) and supervised learning algorithm using linear regression are proposed. To meet the challenges these methods were chosen language PHP, its own established framework CollEntRes, additional tools that helped during the programming process to detect errors and correct the...

This book constitutes the proceedings of the third International Conference on Data Stream and Mining and Processing, DSMP 2020, held in Lviv, Ukraine*, in August 2020. The 36 full papers presented in this volume were carefully reviewed and selected from 134 submissions. The papers are organized in topical sections of hybrid systems of computationa...

Neuro-fuzzy models have a proven record of successful application in finance. Forecasting future values is a crucial element of successful decision making in trading. In this paper, a novel ensemble neuro-fuzzy model is proposed to overcome limitations and improve the previously successfully applied a five-layer multidimensional Gaussian neuro-fuzz...

Data mining and processing of Big Data is key problem in developing intelligent Internet of Things (IoT) applications. In this article, the multidimensional wavelet neuron for pattern recognition tasks is proposed. Also, the learning algorithm based on the quadratic error criterion is synthesized. This approach combines the benefits of the neural n...

The paper proposes a classifying neuro-fuzzy system intended for operating under short training set and nonconvex classes conditions. The proposed system is constructed of feedforward four-layered architecture that solves task of probabilistic classification and fuzzy associative memory which is based on autoassociative evolving memory on fuzzy bas...

The problem of fault detection of time series properties is attractive for many researchers in different areas for a long enough time. The results of the solution of this problem are used in many areas, such as monitoring of the manufacturing processes, control of moving object, bioinformatics, medical diagnostics tasks, and video stream processing...

Time series forecasting can be a complicated problem when the underlying process shows high degree of complex nonlinear behavior. In some domains, such as financial data, processing related time-series jointly can have significant benefits. This paper proposes a novel multivariate hybrid neuro-fuzzy model for forecasting tasks, which is based on an...

In the paper the architecture of Deep Stacking Convex Neuro-Fuzzy System for data stream processing in on-line mode is proposed. The advantage of proposed system is that its layers are formed by multivariate modification of hybrid generalized additive neuro-fuzzy system. Such system is characterized by simplicity of computational implementation, hi...

In the paper the method of fuzzy clustering task for multivariate short time series with unevenly distributed observations is proposed. Proposed method allows to process the time series both in batch mode and sequential on-line mode. In the first case we can use the matrix modification of fuzzy C-means method, and in second case we can use the matr...

In the paper the method of fuzzy clustering task for multivariate short time series with unevenly distributed observations is proposed. Proposed method allows to process the time series both in batch mode and sequential on-line mode. In the first case we can use the matrix modification of fuzzy C-means method, and in second case we can use the matr...

In the paper a two-layer encoder is proposed. The nodes of encoder under consideration are neo-fuzzy neurons, which are characterised by high speed of learning process and effective approximation properties. The proposed architecture of neo-fuzzy encoder has a two-layer bottle neck” structure and its learning algorithm is based on error backpropaga...

In this book chapter, we have considered a topical problem of intelligent energy management, which arises in the context of an intensively developed science direction?Green IT. The hybrid neuro-neo-fuzzy system and its high-speed learning algorithm are proposed. This system can be used for on-line prediction of essentially non-stationary nonlinear...

Techniques to detect the flame at an early stage are necessary in order to prevent the fire and minimize the damage. The flame detection technique based on the physical sensor has limited disadvantages in detecting the fire early. This paper presents the results of using local binary patterns for solving flames detecting problem and proposes modifi...

This paper presents a novel algorithm for detection certain types of emergencies relating to fire, smoke and explosions by processing the data recorded from the camera monitoring, based on cascaded approach. First, the combination of Adaboost and Local binary pattern (LBP) are using for getting Region of Interest (ROI) and reducing time complexity....

The paper describes the image superresolution method with aggregate divergence matrix and automatic detection of crossover. Formulation of the problem, building extreme optimization task and its solution for solving the automation determination of the crossover coefficient is presented. Different ways for building oversampling images algorithms bas...

In the paper the multivariate adaptive hybrid neuro-fuzzy system is proposed that allows to process nonstationary information disturbed by noises in sequential mode and also has smaller number of tuned parameters comparatively with known neuro-fuzzy systems. This proposed system can be used in on-board applications and, first of all, industrial pla...

Здійснено аналітичний огляд сучасних методів детектування вогню за допомогою засобів комп'ютерного зору. Внаслідок проведеної роботи виявлено, що на сьогодні проблема розроблення системи детектування полум'я є актуальною та потребує подальших досліджень. Представлено порівняльний аналіз ефективності роботи окремо взятих методів та відзначено пробле...

The paper describes the method image superresolution from two frames on the basis of aggregate divergence matrix elements of the theory and genetic algorithms. Shows different ways for building oversampling images algorithms based on the proposed method. Experimentally established the effectiveness of the procedures oversampling images at high zoom...

This thesis is devoted to development of visitor's queue density analysis and registration method for a retail videosurveillance systems. Developed method of foreground segmentation is based on initial background modeling, selective temporal median filter and local binary patterns. Based on the literature review, a problem statement has been examin...

In the paper, a new hybrid generalized additive wavelet-neuro-fuzzy-system of computational intelligence and its learning algorithms are proposed. This system combines the advantages of neuro-fuzzy system of Takagi-Sugeno-Kang, wavelet neural networks and generalized additive models of Hastie-Tibshirani. The proposed system has universal approximat...

In this paper we propose architecture of hybrid generalized additive neuro-fuzzy system. Such system is hybrid of the neuro-fuzzy system of Wang-Mendel and the generalized additive models of Hastie-Tibshirani. Proposed hybrid generalized additive neuro-fuzzy system can be used for solving different tasks of computational intelligence and data strea...

One of the primary tasks of videoframes processing of digital videostreams is to quantify the geometric structure of moving objects, in order to extract information from videostreams frames. In this paper we purpose a recursive algorithms that are based on morphological conditional dilation and conditional erosion operations. Practical implementati...

The new learning-based image super-resolution method is described in this article. The process of increasing the resolution of video frames or images from a set according to the method is based on the weight coefficients of synaptic connections. These coefficients are obtained by the learning neural-like structure on a pair of images of low and hig...

This article considers two algorithms which are designed for taking into consideration such events as the light changing in the video frame, the background micro-movements and occlusion of an object. These events determine the negative influences on the segmentation process, because they often arise in the eyeshot of surveillance cameras that are o...

In this paper we propose architecture of hybrid generalized additive neuro-fuzzy system. Such system is hybrid of the neuro-fuzzy system by Wang-Mendel and the generalized additive models by Hastie-Tibshirani. Proposed hybrid generalized additive neuro-fuzzy system can be used for modelling technological process, identification and prediction of no...

Based on the combined usage of developed methods of sustainable segmentation and local magnitudes proposed a new scheme of watermark embedding into audio signals. The proposed scheme of audio signals watermarking ensures the reliability of several types of attacks, such as noise, cropping, resampling, re-quantization, compression and low frequency...

The most common issue of a speech signals analisys and artificial intelligence systems development is determining of temporal and frequency charactristics. That’s because any undetermined signal is defined as a nonlinear object. But it is always possible to select time interval from the signals with a given discretization period. Such an interval i...

Physical process of blurring emergence has been analyzed. It has been proved through conducted experiments that image blurring formation is adequately described by the model based on convolution, i. e. wrapping. It is shown that blurring center or discrete function of point scattering comprises information about trajectory and uniformity of motion,...

The paper suggests a new watermarking scheme based on invariant method of segmentation and the use of local magnitude for marking speech signals. The watermark is embedded in the chosen form at peaks with the spectrum magnitude of each nonoverlapping frame of audio signal.

The new generation mathematical model of single-phase inverter is proposed. The iron-clad transformer is described by equation of quasi stationary electromagnetic field vector potential. The adapter operation of thyristors is described by logical values which reflect the real physical process. Capacitor is described by ordinary differential equatio...

This paper contains one of the approaches of partial motion and out-of-focus image blur detection analysis. Discrete Fourier transform is used to distinguish blurred and unblurred areas of artificially distorted image.

There is a new method to increase the strength of RSA algorithm, using disjunctive partitioning of images.

The use of the method k-means for clustering language signal areas of Polish diphthongs is considered.

There is proposed the mathematical model of calculation of electromagnetic and thermal fields of single phase bridge converter by given voltage. The electric valve action is circumscribed by ideal key method. The spatial-time discretization algebraic equations are solved by the top relaxation method. The results of computation are given.

Based on statistical parameters of the image representation model, interval estimation values for parameters of the model have been suggested, what aimed at one-type image set filtration organisation. We have determined the peculiarities of each representation model usage and given instances of practical experiments.

The offered algorithm stability to illegal decoding fully corresponds to RSA algorithm stability. Nevetherless, according to practical utilization of described RSA modification, in the coded image we managed to achieve complete noising and avoid domains of similarity (Fig. 5). At the same time to display outlines of the sent image becomes practical...

Mathematical models of separate and operative presenting of images and their sets have been outworked. We considered possibility of image presenting through forming diverse image coverings.

There are proposed method of acceleration of large string search in database of learning management system