Search

Genomic prediction models are often calibrated using multi-generation data. Over time, as data accumulates, training data sets become increasingly heterogeneous. Differences in allele frequency and linkage disequilibrium patterns between the training and prediction genotypes may limit prediction accuracy. This leads to the question of whether all available data or a subset of it should be used to calibrate genomic prediction models. Previous research on training set optimization has focused on identifying a subset of the available data that is optimal for a given prediction set. However, this approach does not contemplate the possibility that different training sets may be optimal for different prediction genotypes. To address this problem, we recently introduced a sparse selection index (SSI) that identifies an optimal training set for each individual in a prediction set. Using additive genomic relationships, the SSI can provide increased accuracy relative to genomic-BLUP (GBLUP). Non-parametric genomic models using Gaussian kernels (KBLUP) have, in some cases, yielded higher prediction accuracies than standard additive models. Therefore, here we studied whether combining SSIs and kernel methods could further improve prediction accuracy when training genomic models using multi-generation data. Using four years of doubled haploid maize data from the International Maize and Wheat Improvement Center (CIMMYT), we found that when predicting grain yield the KBLUP outperformed the GBLUP, and that using SSI with additive relationships (GSSI) lead to 5–17% increases in accuracy, relative to the GBLUP. However, differences in prediction accuracy between the KBLUP and the kernel-based SSI were smaller and not always significant.

Circular dichroism (CD) spectroscopy is highly sensitive to the secondary structure (SS) composition of proteins. Several methods exist to either estimate the SS composition of a protein or to validate existing structural models using its CD spectrum. The accuracy and precision of these methods depend on the quality of both the measured CD spectrum and the used reference structure. Using a large reference protein set with high-quality CD spectra and synthetic data derived from this set, we quantified deviations from both ideal spectra and reference structures due to experimental limitations. We also determined the impact of these deviations on SS estimation, CD prediction, and SS validation methods of the SESCA analysis package. With regard to the CD spectra, our results suggest intensity scaling errors and non-SS contributions as the main causes of inaccuracies. These factors also can lead to overestimated model errors during validation. The errors of the used reference structures combine non-additively with errors caused by the CD spectrum, which increases the uncertainty of model validation. We have further shown that the effects of scaling errors in the CD spectrum can be nearly eliminated by appropriate re-scaling, and that the accuracy of model validation methods can be improved by accounting for typical non-SS contributions. These improvements have now been implemented within the SESCA package and are available at: https://www.mpibpc.mpg.de/sesca.

Reaction-based de novo design refers to the in-silico generation of novel chemical structures by combining reagents using structural transformations derived from known reactions. The driver for using reaction-based transformations is to increase the likelihood of the designed molecules being synthetically accessible. We have previously described a reaction-based de novo design method based on reaction vectors which are transformation rules that are encoded automatically from reaction databases. A limitation of reaction vectors is that they account for structural changes that occur at the core of a reaction only, and they do not consider the presence of competing functionalities that can compromise the reaction outcome. Here, we present the development of a Reaction Class Recommender to enhance the reaction vector framework. The recommender is intended to be used as a filter on the reaction vectors that are applied during de novo design to reduce the combinatorial explosion of in-silico molecules produced while limiting the generated structures to those which are most likely to be synthesisable. The recommender has been validated using an external data set extracted from the recent medicinal chemistry literature and in two simulated de novo design experiments. Results suggest that the use of the recommender drastically reduces the number of solutions explored by the algorithm while preserving the chance of finding relevant solutions and increasing the global synthetic accessibility of the designed molecules.

Feature detection and matching are an essential component of many computer vision applications. Consider the two pairs of images shown in Figure 4.2. For the first pair, we may wish to align the two images so that they can be seamlessly stitched into a composite mosaic (Chapter 9). For the second pair, we may wish to establish a dense set of correspondences so that a 3D model can be constructed or an in-between view can be generated (Chapter 11). In either case, what kinds of features should you detect and then match in order to establish such an alignment or set of correspondences? Think about this for a few moments before reading on.

This chapter introduces the reader to the fundamentals of descriptive statistics and data visualisation. Descriptive statistics focus on the development of methods for describing a given data set. They can be divided into two main groups: measures of central tendency, which examine the average behaviour of the data set, and measures of dispersion, which examine the spread of the data set. Measures of central tendency, such as the mean, mode, and median, are introduced, while measures of dispersion considered include the range, standard deviation, variance, median absolute difference, and skew. Also, quantiles and outliers are introduced as ways to describe a data set. Data visualisation focuses on developing a set of rules for effectively displaying data visually. Common data visualisation methods such as bar charts, histograms, pie charts, line charts, time series plots, box-and-whisker plots, scatter plots, probability plots, tables, and sparkplots are explained with detailed examples and methods of construction. The different approaches are illustrated with suitable examples, including a comprehensive analysis of a data set from a friction factor experiment. By the end of this chapter, the reader should be able to apply the principles of data description and visualisation to meaningfully portray the key properties of a given data set.

A social accounting matrix (SAM) is shown to represent the ideal set of data to endogenize household consumption, as it contains a full description of the generation, redistribution and spending of income. Type II multipliers and interregional spillovers of an interregional SAM model are both larger than those of the standard, Type I input–output model, whereas exogenous final demand is smaller. Type II multipliers are shown to represent an upper limit for the true multipliers. Type III multipliers are smaller, as intensive income growth of existing jobs needs to be multiplied with smaller marginal instead of average consumption/output ratios. Type IV multipliers are even smaller, as they include the negative feedback of increasing employment on unemployment benefits. Endogenizing remaining final demand leads to ever larger, less plausible multipliers.

The main features of the 2020 version of global Numerical Weather Prediction (NWP) model ARPEGE run operationally at Meteo-France are described. This spectral model, developed in collaboration with the Integrated Forecasting System (IFS) of ECMWF, has a tilted and rotated horizontal grid that allows to reach a resolution of 5 km over Europe. The initial conditions are provided by an incremental 4D-Var data assimilation system with a 6-hour time window. Two inner-loops are performed respectively at 100 and 40 km. A comprehensive set of observations is assimilated with a dominance of satellite data representing 90% of them. However in terms of information content, conventional observations reach a fractional value of 20%. A 50-member ensemble data assimilation system based on low resolution 4D-Var is used to estimate daily background error covariances. The most recent improvements on this system regarding model resolutions, ensemble size and observation usage, that took place between mid-2019 and mid-2020, are presented with a selection of evaluations in terms of analysis and forecast skill scores.

Globally, tuberculosis (TB) is the leading infectious killer in the world before pandemia. This paper presents the result of optimizing convolutional neural network architecture for the detection of acid-fast stained TB bacillus. The experimental set contains the segmentation results of microscopy images of the patients sputum stained by the Ziehl–Neelsen method. The authors constructed an experimental algorithm for optimizing the original convolutional neural network model, including optimizing the model dimension, data augmentation, adjusting the model parameters, and improving regularization. The authors built few models of convolutional neural networks (CNN) models to recognize TB bacillus, which showed the maximum value of metrics in the experiment.

Inverse kinematic control of industrial robotic manipulators is extensively used, for this reason it is necessary to improve current direct and inverse kinematic solutions. While some previous solutions present partial sets of equations and others have some inconsistencies, this study presents the complete derivation of an alternative set of equations for the inverse kinematic solution for Universal Robots robotic arms, particularly the UR5. The herein inverse kinematic solution is obtained by applying the conventional Denavit-Hartenberg algebraic method and is validated with the direct kinematic solution.

In this chapter, we first give the general and abstract definition of a deterministic dynamical system. This definition has as its basic ingredients a state space whose points characterize the state of the system and a family of maps that determines the time evolution of states according to a dynamical law. The state space is comprised of the minimal set of variables that determines all properties of interest. To qualify as a dynamical system the family of maps must satisfy a composition rule to conform with the additive structure of time.

In the presented article, the authors investigate the issue of the role and significance in the application of technologies for processing large data arrays using Big Data in the development of the pharmaceutical industry. Taking into account all the complexity and knowledge-intensiveness of the process of developing and launching a new drug on the market, the introduction of new approaches to the collection of additional information can significantly facilitate and simplify this procedure. It is also important that the use of Big Data technologies in pharmaceuticals significantly reduces the cost of the development process and minimizes possible risks associated with human health. In this regard, the relevance of the application of Big Data technology in the pharmaceutical business today is extremely relevant, since both manufacturing companies and consumers benefit. Within the framework of this work, the authors set the goal to analyze the main areas of the pharmaceutical business, the use of Big Data technology in which today is the most in demand.

In this paper, iterative procedure for partial fraction approximation of vector functions in Laplace domain is presented. First, the essentials of the original formulation of vector fitting methodology for strictly proper rational approximations are outlined. Then a modification is presented for more efficient application to vector functions. Proposed modification reduces the problem of fitting all elements of vector function to fitting only some of them and sum of the rest. After the set of common poles is obtained, the residues are calculated for all functions as usual. Next, the algorithm for adaptive rational approximation of vector functions is thoroughly presented. Proposed technique consists of successive determination of partial fraction approximations with increasing orders until the specified convergence criterion is satisfied. Detailed results of the numerical example are provided to assess accuracy and efficiency of the proposed procedure.

We developed an automatic framework to classify sperm heads as normal or abnormal using image processing and machine learning techniques. The framework segments each sperm head using a color-space-based classification method. A novel set of morphological features is proposed to better describe the sperm head morphology. Finally, a supervised learning model is trained and tested to analyze the feature data for classification. To train and test the model, a publicly available dataset of human sperm heads was used. All sperm samples were manually labeled as normal or abnormal according to the strict criteria of the World Health Organization laboratory manual for the examination and processing of human semen. The segmentation method preserves shape without losing key morphological aspects. The classification model based on morphological descriptors produces better discrimination as compared with the traditional shape descriptors, achieving a 92% accuracy in the discrimination of normal or abnormal spermatozoa.

In the Best-Worst Method (BWM), the criteria weights are typically characterized by intervals, each value in the interval representing an optimal weight for the associated criterion according to the preferences of the decision-maker. While the intervals can potentially provide the DM with more information, it makes it challenging to process the weights of a group of DMs, e.g., for computing the aggregated weights or evaluating a set of alternatives with respect to the interval weights. The problem compounds when the performance matrix of the alternatives is also acquired by the BWM. This paper presents a Monte Carlo approach to address these shortcomings. First, a Monte Carlo approach is developed to compute the aggregated interval weights for group decision-making problems, as well as the extent to which a criterion is more important than another based on the preferences of the group. Second, another Monte Carlo method is developed to evaluate and compare the alternatives based on the interval weights. The experiments validate the applicability of the proposed approach for the BWM.

The article discusses the construction of a model of information support for the process of the quality management system, and its integration into a single information space, which separates all stages of the product life cycle. Also, the article presents an algorithm for determining vector estimates that should be presented to the information and analytical center, in order to ensure the division of the set of alternatives into classes of solutions for a relatively small number of calls. The application of this algorithm will allow to effectively solve the problems of improving the QMS, by means of automatic classification of the decisions made in order to minimize the risk of not achieving the planned quality indicators at one or another level or stage of the product life cycle.

An improved real-time multi-target tracking algorithm based on the combination of improved YOLOV3-Tiny and deep-sort is proposed. The target tracking environment and target are simple, so it is easier to use the self-built data set to fit and test the vehicle it belongs to. In order to prevent the phenomenon of over-fitting, the affine transformation algorithm is used to expand and transform the data set of heavy trucks. After adding the self-built data set, it has a better effect on the original data set test. By replacing the main network of Yolov3-Tiny with a more concise neural network structure and using the effective combination of Kalman filter and Hungarian algorithm, the whole structure is more portable and fast, easier to deploy to small computers even mobile terminals. Compared with Faster RCNN, MobilenetV2, YOLOV3-Tiny and other algorithms on self-built data sets and Pcal VOC data sets, the proposed algorithm has higher real-time performance and accuracy, and robustness. After the neural network is improved, the accuracy of the proposed algorithm is improved by more than 2% compared with the original algorithm. And the computing speed is faster, and the network is more portable.

The goal of this research is to estimate the degree to which a person appreciates a given piece of music, predict and recommend the next music piece/song. Listening to music is subjective and dynamic. Such profiles may be helpful in applications like digital forensics. For example, prisoners, people with depression or other health issues. This research work involves designing a deep learning algorithm to extract attributes of music, such as tempo, and pitch, etc. A set of such attributes is identified using attribute classification stage, to define the liking index (Li) based on the user responses, subsequently appearing as inputs to the neural network. The learning capability of the neural networks is employed to build the music profiles in response to the inputs. The attribute classification stage involves clustering using the K-means algorithm for the created music profiles. Synthetic data (user-selected set of music pieces) was created initially for testing the proof of concept. Preliminary results show that for K = 5, the classifier results show minimum outliers with well-defined clusters. Real data and steps to validate the proposed algorithm are the next steps to be carried out. After validation, the model will be tested with user data to confirm the design and its effectiveness.

The article discusses the features of electronic document management of technical documentation of railway automation and telemechanics. The article deals with the problems of synthesizing the mathematical description of the electronic document management of technical documentation for railway automation and telemechanics. For this, an overview of the actual processes of making, checking and applying technical documentation was carried out on the example of maintaining custom specifications for automation and telemechanics systems, which made it possible to identify document flow scenarios, as well as protocols of technical documentation properties. A method for creating a model of electronic document management of technical documentation based on the apparatus of graph theory has been advanced. Methods for determining sets for the proposed model are described, an algebra of workflow of TD using graphs is proposed. To set the matrix type of document flow representation, it is proposed to use a set of flat rectangular matrices, each of which represents the state of the system in some discrete unit of time. The resulting matrices of incidents determine the graph model of the document flow of the process under consideration, and the set of these matrices sets all possible scenarios for the movement of documents in the process, describes all possible states of documents and determines possible participants. The use of the proposed method for constructing a mathematical model based on graph theory makes it possible to develop reliable applied software for solving problems of operational document flow of technical documentation.

Flexible and varied manual assembly processes in the automotive industry are often based on manual labor. While simulation can be used to improve planning to maximize efficiency while minimizing ergonomic issues for workers, common simulation tools require extensive modeling time. In such simulations, the users are often process engineers who want to easily create complex human motion simulations. This paper presents a concept developed to create complex human motions for interacting with objects in a production environment with little effort. The concept separates between geometric constraints and the semantic meaning of the respective geometry. With a set of data types developed for this purpose based on a unified ontology, a range of geometric and semantic information can be specified for arbitrary objects. In this way, an action-specific motion generator can be used to define the appropriate motion for the interaction with an object depending on the action without defining case-specific constraints. For a first proof, the concept is tested and demonstrated in the assembly of a pedal car and sitting on a chair at a manual workstation. Based on the use case, the effect of effort reduction is shown.

Detecting faults in manufacturing applications can be difficult, especially if each fault model is to be engineered by hand. Data-driven approaches, using Machine Learning (ML) for detecting faults have recently gained increasing interest, where a ML model can be trained on a set of data from a manufacturing process. In this paper, we present a use case of using ML models for detecting faults during automated screwdriving operations, and introduce a new dataset containing fully monitored and registered data from a Universal Robot and OnRobot screwdriver during both normal and anomalous operations. We illustrate, with the use of two time-series ML models, how to detect faults in an automated screwdriving application.

Model estimation is an important issue in signal processing since it enables to characterize the signal and to obtain some parameters providing information about its shape. An example of a model widely used in various fields of science is the sum of exponentially damped sinusoids. In this work, a new method for parameters estimation of such a model is proposed. This method uses nonlinear least squares together with Fourier transform to estimate parameters of the sum of damped sinusoids. For model order selection, a bootstrap method is utilized. First, the presented method is verified on simulated signals. The efficacy of model order selection is assessed for signals with a different set value of signal-to-noise (SNR) ratio. It is shown that for the models with orders from 2 to 7, the order was estimated correctly in all trials for SNR greater than 20 dB. If a proper model order is selected, the set values of signal parameters are included in the confidence intervals for parameters values, determined by the nonlinear least squares method. Afterward, the described method is used for modeling the event-related potential in magnetoencephalography (MEG) signal. The method gives good fitting results with \(\text {R}^{2}\) values greater than 0.9, showing its potential in MEG signals modeling and the assessment of brain activity.

In the CBAC (Capability-Based Access Control) model for the IoT (Internet of Things), subjects are issued capability tokens, i.e. a set of access rights on objects in devices, by device owners. Objects are data resource in a device which are used to store sensor data and action data. Through manipulating objects of devices, data are exchanged among subjects and objects. Here, the illegal information flow and the late information flow occur and subjects can get data which the subjects are not allowed to get. In our previous studies, protocols are implemented to interrupt operations implying both illegal and late types of information flows. Here, the request processing time takes longer as the number of capability tokens whose signatures are verified in devices increases. Hence, an MRCTSD (Minimum Required Capability Token Selection for Devices) algorithm to reduce the number of capability tokens used is proposed. However, the more number of capability tokens are sent from subjects, the more complex the capability token selections are. Since the devices support just low processing power and smaller size of memory, it is important to avoid concentrating loads in devices. In this paper, MRCTSS (MRCTS for Subjects) algorithm where subjects selects minimum required capability tokens and sends them to devices is proposed. In the evaluation, it is shown that the size of a UDP datagram in a get access request is made smaller by the MRCTSS algorithm.

In recent years, along with the increasing sophistication of mobile terminals, systems have been produced to estimate a user’s position using global positioning system (GPS) technology of smartphones and tablet terminals. Some systems use augmented reality (AR) to present a route to the user destination. However, most such systems only show a direction of travel and signs using AR, lack affinity from the viewpoint of humans, and have some points for which GPS cannot be positioned accurately indoors. Therefore, for this study, we developed a mixed reality (MR) navigation system that can set navigation indoors freely without depending on GPS. The system can also present intuitive routes.

In online machine learning systems, a computer gathers sets of training data from some data sources and trains a machine learning model every it gets a set. In the cases that the time required to gather a set is long such as the case of the data gathering from connected vehicles, the delay for reflecting the observed environmental values included in the training data to the model lengthens. In this paper, we propose a system to reduce the delay for reflecting observed environmental values to the models suppressing the increase of the validation loss. Our proposed system cyclically broadcasts the parameters of the machine learning model to the data sources and the data sources calculate the result of the loss function for their observed training data. We evaluated the proposed system assuming a machine learning system for connected vehicles. The vehicles of that training data give a larger value to the result than a given threshold send the training data to the computer for training the machine learning model. Our experimental evaluation revealed that our proposed methods can achieve lower validation loss values than a conventional method.

This paper addresses the balancing of a haptic device by using a set of springs. The proposed haptic device is based on a purely translational parallel manipulator. Given the specific application as haptic device, it is necessary to eliminate or minimize the effect of gravity for achieving a suitable positioning and force feedback. Given the peculiarities of this haptic device, this paper reports the design, construction and preliminary testing of a simple and low-cost balancing solution based on an external set of springs. Simulation results and preliminary tests are reported for showing the engineering feasibility and effectiveness of the proposed solution.

This paper presents a method of building a 3D model of the geographic location described in Tana French’s novel Faithful Place. Its action is set in Faithful Place, an invented Dublin working-class neighbourhood, based on the street existing in the past and moved to a different area of the city. It can be defined as M. Foucault’s heterotopia and is central to the novel. The work uses the methodology of literary cartography, an interdisciplinary research area developing verifiable techniques for mapping literary spaces, complementing it with 3D modelling. The analysis is performed in three steps. First, the prototype of the fictitious street is identified using the historic map of Dublin. Then, all geographic information from the novel is put on the map to make a conclusion about the most probable location of fictional Faithful Place. Last, a 3D model of the location is built that generates new knowledge about the literary space featured in Faithful Place and enhances the message of the novel.

The aim of our work is to develop a software that could detect toxicity in the Russian segment of social media. In this paper, we investigated the problem of toxic detection in messages in Russian language. We implemented a set of features using selected vector models, trained some classifiers on the dataset about fourteen thousand annotated messages and compare results. Experiments were conducted with a calculation of accuracy, precision, and recall values. F1 measure reached the value 0.91, accuracy value is 0.87.

Monitoring of acid mine drainage is an important ecological task. Remote sensing provides means of accurate operational observations of the Earth in a wide range of spectrum. Numerous remote indices of surface water state/quality have been developed based on remote sensing data. Monitoring of small-sized objects requires a fine spatial resolution which generally implies a coarser spectral resolution. Hence, only a qualitative knowledge of water optical features can be taken into account which makes it difficult to perform a quantitative analysis of water compounds. A possible approach is a comprehensive analysis of a numerous indicators, which can be sensitive not only to variations of water optical properties due to pollution but also to changes of the course of biological and sedimentation processes caused by it within next days. The aim of this work is selection and/or construction of a set of such indicators and presents the initial stage of the investigation. A list of used remote indicators (around 20 in total) is formed; an approach to their qualitative comparison is introduced and tested; the prospects of the future investigation are discussed; and some recommendations are given.

The decomposition of HD-EMG into motor unit (MU) discharge timings permit a detailed window into the motoneuronal manifestation of motor intent. Recently, the feasibility of MU-driven wrist joint angle estimation was preliminarily demonstrated although the influences of certain parameter selections have yet to be fully investigated. Here, a decomposition algorithm was used to predict wrist joint kinematics over three DoFs in a pseudo-online manner. Three separate estimator types were tested and the effects of two key parameters on their prediction accuracies were studied: the decomposition extension factor and process window length. Pre-recorded EMG from four able-bodied subjects was decomposed in a simulated real-time manner as to permit parameter scanning, with the tested estimators being linear regression (LR), linear discriminant analysis (LDA), and LDA with LR for proportionality control (LDA-LR). Results showed the best performing combination of parameters were an extension factor of 8 with window length of 50 ms which allowed the LDA-LR estimator to yield an average R2 value of 0.86 ± 0.05. Under the most computationally demanding set of parameters, the median processing time of the algorithm on a desktop computer was 47 ms which was within the update rate of the proposed system. Such results also indicate that parameters optimal for online control applications deviate from those ideal for offline physiological studies.

The open-source software (OSS) market comprises of thousands of products and applications with different quality. The primary concern of the individuals and organizations is to evaluate the quality OSS products and packages. In this chapter, we have demonstrated a MCDM based model to assess the quality of OSS by taking performance and cost-based criteria related with OSS to analyze its quality on the basis of feedback gathered from the users and the experts. To avoid the uncertainty attached with the opinion of the expert, the Maximum-Entropy-Minimum-Variance-Ordered-Weighted-Aggregation (MEMV-OWA) operator has been incorporated. The criteria weights are calculated by solving a non-linear multi-objective programming problem proposed in the MEMV-OWA operator. The picture fuzzy set information has been used which is an addendum of intuitionistic fuzzy set, representing the human opinion more precisely. In this chapter, we have proposed a model named MEMV-OWA-PF-TOPSIS (Maximum Entropy Minimum Variance-Ordered Weighted Aggregation-Picture-Fuzzy-TOPSIS). A step by step procedure has been exhibited to show its implementation in real-life problems. A numerical illustration related to the software quality assessment of OSS on the basis of their performance and cost-based criteria has also been provided in this study.

Consider the set of inferences that are acceptable to use in all our theory building endeavors. Call this set of inferences the universal theory building toolkit, or just ’the toolkit’ for short. It is clear that the toolkit is tightly connected to logic in a variety of ways. Beall, for example, has argued that logic just is the toolkit. This paper avoids making a stand on that issue and instead investigates reasons for thinking that, logic or not, the toolkit is substructural. It is presented as a dialogue for the simple reason that it summarizes a range of dialogues on this subject that the author has had with various folks over the past few years. The method I use to investigate the toolkit is inspired in both philosophical and technical details by Alasdair Urquhart’s work on semantics for relevance logics from the early 1970s.

The operational semantics of Urquhart is a deep and important part of the development of relevant logics. In this paper, I present an overview of work on Urquhart’s operational semantics. I then present the basics of collection frames. Finally, I show how one kind of collection frame, namely, functional set frames, is equivalent to Urquhart’s semilattice semantics.

This chapter discusses simulation results of MPC algorithms based on Wiener models applied to the neutralisation reactor. At first, the process is shortly described and identification of the Wiener model is discussed. Polynomials and neural networks are used in the nonlinear static block of the model. Effectiveness of both model classes is compared. Implementation details of different MPC algorithms are given. Next, MPC algorithms are compared in terms of control quality and computational time in the classical set-point following task and, additionally, some constraints are imposed on the predicted value of the controlled variable.

Smart buildings have grown in numbers and capabilities worldwide. Smart buildings rely on the sensor networks and Internet of Things (IoT), which allows the development and optimization performance through various methods, including efficient energy management solutions and predictive technologies for future decisions. Smart buildings generate data in real-time, and human-based choices are often taken in real-time. IoT provides connectivity for the various sensors. Smart buildings must ensure the proper operation with minimum time and energy with regards to dynamic human demands and performance in real-time. Blockchain technology can ensure data integrity and regulated autonomous flow of information between users who have a varying degree of trust. Therefore, the combination of Blockchain and the IoT can achieve data integrity between IoT devices installed in the smart buildings where multiple humans are involved and have different non-overlapping expectations from the building’s environment and response. In this paper, a set of predictive technologies are analyzed, and their common parameters are shown to affect the possible prediction and visualization of the IoT data for users. A digital twin based on Blockchain is proposed where individual users owing some sensors can use and visualize processed data based on rules set by the owners in the Blockchain.

The bisection method is one of the most used methods of solving non-linear equations, it is based on existence of a solution (root) in a given interval on a real number line, the interval is divided into two equal parts, it can be obviously seen that the root will be contained by one of the intervals, the interval that contained the root is later divided into another two sub-intervals while the interval that does not contain the root is dropped and never used again. The processes continued until the desired stopping criteria is reached, the interval that contain the roots is reduced to set of points, the root and its neighborhood. The root and its neighborhood may approximately serve as the desired root. In this research a new method, a modification of bisection method namely the trisection method and pentasection method are introduced, this method will be tested using several selected functions, the obtained results are then analyzed based on the number of iterations, the CPU time. The result from this research showed that the higher the number of the divisions of the interval, the higher the CPU time, while the number of iterations that yielded the result (root) reduces drastically. Moreover, the trisection method works faster than the Pentasection method.

The conducted research belongs to the scientific field of data mining in relation to the problems of classifying the state of objects and systems. A method for diagnosing the state of a complex chemical-energy-technological system for the production of phosphorus from wastes of apatite-nepheline ores is presented. The method is based on the use of an ensemble of deep neural networks, consisting of a parallel working recurrent network of long short-term memory and a one-dimensional convolutional neural network. These networks solve the problem of multi-class unambiguous classification on the same set of diagnostic data. The results of assessing the state of the technological system by the recurrent and convolutional neural networks are compared in the block for comparing the classification results, and if they match, the final state assessment is formed, otherwise, a warning message is issued about the need for a person (operator) to participate in making a decision. The scientific novelty of the proposed method lies in the use of comparing the results of data processing by neural networks of different types to obtain the final classification results. The study was carried out as part of the development of a digital twin of the technological system. The choice of the apparatus of deep neural networks makes it possible to implement the self-specification procedure of the digital twin at the stage of the system operation. The software implementation of the proposed method to diagnose the state is made in Python 3.6 using the TensorFlow machine learning libraries and the Keras framework. The results of the program are given, which showed that duplication of the processing of the same data by neural networks of different architectures provides greater confidence in the results of classification or contributes to the decision to repeat it.

This paper presents an approach to determining the psychological characteristics of a user of social networks by analyzing text messages on social networks. The approach is to classify the user's copyright texts using machine learning. Parsing the extracted text data suggests using the vectorizer - transformer - classifier sequence. The training dataset included the results of a Big Five analysis of user surveys and a set of copyright text data from social media pages. The effectiveness of the proposed approach was confirmed by the results of experiments with open data of more than 1000 users of the social network Facebook.

The aim of the work is to consider the problems of calculating and design of reinforced concrete columns of circular cross-section, the development of proposals for a simplified analytical calculation of such columns. Studying these problems was carried out by methods of analysis, comparison, description, generalization; the calculation of such columns was carried out by the numerical method using the LIRA software package, by the analytical method using the method of simplifying the mathematical model (algorithm), as well as the graphical method. The investigation showed the presence of problems in the design of reinforced concrete columns of circular cross-section, made of both pipe-concrete and having a solid circular cross-section. The set of rules for the design of steel-reinforced concrete structures, introduced in 2016, contains only general recommendations for the design, in which issues such as the joint operation of the shell-pipe and the concrete core during eccentric compression of the concrete are not considered. The existing methods do not fully take into account the properties of building materials, the real stress-strain state of the concrete core and the steel shell, the peculiarities of their joint work. There are also difficulties in analytical calculation of solid circular columns according to the algorithm, recommended by building codes, which complicates the verification of numerical and analytical calculation methods by comparing them with experimental results. The article presents a simplified graphical method, developed by the authors, for calculating columns of solid circular cross-section, based on the algorithm, recommended by the norms.

Reinforced concrete matrices are used in the method of hydro explosion stamping for the manufacture of sheet metal products. Reinforced concrete matrices have a number of advantages that ensure the effectiveness of their use. However, there is a problem – low durability under repeated explosive loads. The analysis of scientific publications makes it possible to set the goal and objectives of the research. Thus, the paper analyses ways of improving the constructive of reinforced concrete ring-shaped matrices for hydro explosion stamping and methods of calculating matrices. The author describes the developed design of the matrix, to which he additionally install a compensation layer and a hydraulic chamber. The paper further describes the developed method of calculating ring-shaped reinforced concrete matrices. The author's proposed method for calculating strength and deformation is based on the method of finite integral transformations. A theoretical calculation was performed using various mathematical models. Experimental researches of a series of matrices samples under explosive loads have been carried out. Comparison of theoretical and experimental results is made. Moreover, both our own research and other authors. As a result, a calculation method was obtained that allows one to determine the values of stresses and deformations in the matrix under the action of impulse loads at any time and in any part of the matrix.

The railway is a strategic part of the industrial infrastructure of many countries. The overall quality of the railway track structure is affected individually and in various combinations of geometric and material parameters in the form of an extensive set of quality indicators. The quality of a railway track structure during its service life goes through several phases of utilisation (quality degradation) and repairs (quality rehabilitation). Quality indicators represent parameters and factors that are in individual technical groups of the structure capable to objectively describe the quality of their condition and behaviour. At the border of the degradation and rehabilitation phases, the railway manager decides on the implementation of the optimal technical, technological, time and economic format of quality rehabilitation. The decision-making process is subject to the structure and result of the evaluation of the diagnostic data. Although various factors of mechanical properties of the track structure have been investigated for its operational quality in the recent period, the quality of track geometry represented by geometric parameters remains the focus in the decision-making process on the rehabilitation of track quality. This article examines the quality of a specific test section of a railway track from the point of view of the quality of the track geometry. The structure of diagnostic data, the method of their collection, determination of quality indicators and their evaluation are described. The evaluation of quality indicators is connected with models for predicting the future state of track geometry quality using regression and correlation analysis of the dependence of the development of determining quality indicators on time.

In the chapter, a posteriori methods for solving multi-criteria synthesis problems are analyzed and modified. An example of their use in modeling and selecting the values of the parameters of the kinematic model of the configuration of a four-fingered robotic gripper for harvesting tomatoes is described. The results of the simulation were compared using the NSGA-II, MOGWO, and MOPSO methods. Variants of the choice of the objective function, weight coefficients and their influence on the set of optimal grip sizes are discussed.

Wireless networks are the foundation of modern communication. This also leads to network data hijacking and network attacks. Despite the fact that different databases and methods exist to combat this, the network intrusion problem is still on the rise. To address this, the proposed research work introduces a faster and better Intrusion Detection System trained on the AWID data-set with faster learning and better accuracy. A unique supervised studying device has been evolved to categorise the community wi-fi as malicious or benign. To discover the wi-fi version by considering the detection wi-fi rate, a combination of supervised gaining knowledge of a set of rules and feature choice technique are used. Through this it is observed that, Artificial Neural Network (ANN) is primarily based on the machine studying with wrapper characteristic selection that outperform support vector machine (SVM) technique even as classifying community traffic. To assess the performance, NSL-KDD dataset with feature wrapping is used to categorise community traffic data with the use of SVM and ANN supervised machine gaining knowledge technique. Comparative studies suggests that, the proposed model is faster and more efficient than different existing fashions to recognize the intrusion detection rate. This paper aims to categorize the threats to Institute of electrical and Electronics Engineers (IEEE) 802.11n well known protocol and classify the contemporary answers and countermeasures to them. In addition, this paper proposes a more accurate, and faster widespread intrusion detection system as a proposed solution for IEEE 802.11 threats.

Gentiana is an important genus of around 360 medicinally important species, majority of which are not well characterized. Despite its importance, very few genomic resources are available for Gentiana L. Till date, the number of informative and robust simple sequence repeat (SSR)-based markers is limited and very few efforts have been made for their development. A set of robust, freely accessible and informative SSR markers for Gentiana is a pre-requisite for any molecular systematic as well as improvement studies in this group of pharmacologically valuable plants. In view of the importance of these plants, Expressed Sequence Tag (EST) sequences of 18 Gentiana species were surveyed for the development of a large set of non-redundant SSR markers. A total of 5808 perfect SSR with an average length of 17 bp and relative abundance of 214 loci/Mb were identified in the analysed 47,487 EST sequences using Krait software. Mapping of the ESTs resulted in gene ontology annotations of 49.14% of the sequences. Based on these perfect SSRs, 2902 primer pairs were designed, and 60 markers were randomly selected and validated on a set of Gentiana kurroo Royle accessions. Among the screened markers, 39 (65%) were found to be cross-species transferable. This is the first report of the largest set of functional, novel genic SSR markers in Gentiana, which will be a valuable resource for future characterization, genotype identification, conservation and genomic studies in the various species of this group of important medicinal plants. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s13205-021-02969-4.

As a result of the research, a combined neural network method, models and modeling tools for the operational control of complex technical systems have been developed and investigated. The analysis of existing methods, models and software for operational management of complex technical systems; the requirements for the developed methods, models and software for managing these systems have been substantiated. A combined neural network method for modeling complex technical systems based on a hierarchical neural network model is proposed, combining the capabilities of analytical and neural network approaches and allowing to significantly increase the efficiency of operational management (quality and efficiency of management decisions), as well as the flexibility of modeling and the possibility of operational management of calculation of structural and parametric optimization of the model based on information about the system obtained in real time. A hierarchical neural network model for the operational control of complex technical systems has been developed, consisting at the lower level of a set of interconnected neural network models corresponding to the elements of the modeled system, and a supervisor at the upper level designed to identify structural errors at the lower level of the model. A method of decision support for managing complex technical systems in real and pseudo-real time, based on the proposed hierarchical neural network model, is proposed. Software tools for combined neural network modeling and decision support have been developed for the operational control of complex technical systems.

Smart contracts had a very fast increasing development in the last years. Once a smart contract is deployed on a blockchain due to code immutability, its residual vulnerabilities cannot be patched. Reducing the number of residual vulnerabilities becomes thus very important and normally is achieved through static analyzers. This paper investigates the physical position (location) of vulnerabilities in Solidity smart contracts. To this purpose, we use a language-independent systematization of vulnerabilities and we consider the outputs of a set of static analyzers processing a representative set of smart contracts. We analyze the distributions of the locations where tools find positive outcomes. We create the ground truth of vulnerabilities for a subset S of smart contracts through manual inspection and we first perform a comparison of the distributions within this set. Then we generalize our findings by comparing the distributions between the manually inspected subset and the full set. Such comparison allows us to identify where certain classes of vulnerabilities are located, suggesting specific areas in Solidity smart contracts where the search for vulnerabilities should focus.

Smart contracts certainly provide a powerful functional surplus for maintaining the consistency of transactions in applications governed by blockchain technology. However, the intended level of automation might cause cascading effects that have to be checked by formal methods of algorithmic proof. Our smart contract formal model analysis framework uses the Finite State Machine (FSM) theory which is a model of behavior composed of states, transitions and actions. In this model, a state stores information about the past, a transition indicates a state change and is described by a condition that would need to be fulfilled to enable the transition. An action is a description of an activity that is to be performed at a given moment. These conditions are properties that are checked during program execution. This formal analysis framework allows us to define a set of invariants on Finite State Machine behavior model and to propose an anomaly detection system based on the invariants of the smart contract.

In this research, the recent development of three dimensional fuzzy sets namely, spherical fuzzy sets is extended to complex spherical fuzzy sets (CSFSs) and the novelty of CSFSs with their operations and properties are investigated. The range of CSFSs membership functions lie in complex plane, not limited to [0, 1]. So, the CSFSs describe complex membership functions in a set in terms of complex numbers which leads to a new mathematical framework and comprehensive study of their properties. Some of the aggregation operators and distance measures are established with their mathematical proof. Further an intelligent system based on EDAS (Evaluation based on Distance from Average Solution) method has been developed using the proposed complex spherical fuzzy aggregation operators. An application is proposed for aviation industry 4.0 in order to assess the performance of catering services. Finally a numerical illustration has been demonstrated for the proposed innovative method.