Computational Science and Its Applications – ICCSA 2022 Workshops

This article presents the implementation of an application developed under the android operating system to evaluate the actions performed after a flight mission in manned aircrafts. The informatics application will be installed on an Android device that will be transported by a crew member of the aircraft. In order to obtain the required signals for the post-flight analysis, two sensors are used: Global Positioning System (GPS) and Inertial Measurement Unit (IMU), which are immersed in the Android device. The information generated at aircraft corresponding to the attitude and position is stored within a SD memory of the android device. The stored data corresponding to the position are: latitude, length, height, velocity, and hour; while the data corresponding to the attitude are Roll, Pitch, and Yaw.

Companies’ websites are vulnerable to privacy attacks that can compromise the confidentiality of data which, particularly in sensitive use cases like personal data, financial transaction details, medical diagnosis, could be detrimental and unethical. The noncompliance of companies with privacy policies requirements as stipulated by the various Data Protection Regulations has raised lot of concerns for users and other practitioners. To address this issue, previous research developed a model using conventional algorithms such as Neural Network (NN), Logistic Regression (LR) and Support Vector Machine (SVM) to evaluate the levels of compliance of companies to general data protection regulations. However, the research performance shows to be unsatisfactory as the model’s performance across the selected core requirements of the legislation attained F1-score of between 0.52–0.71. This paper improved this model’s performance by using Natural Language Processing (NLP) and Deep Learning (DL) techniques. This was done by evaluating the same dataset used by the previous researcher to train the proposed model. The overall results show that LSTM outperform both GRU and CNN models in terms of F1-score and accuracy. This research paper is to assist the Supervisory Authority and other practitioners to better determine the state of companies’ privacy policies compliance with the relevant data protection regulations.

Within the scope of this study, we developed a product recommendation methodology for customers by analyzing shopping behaviors based on user-system interaction data collected on Casper Computer Systems’ website. To achieve the “right product to the right customer” objective, we predict customer interests using a collaborative filtering algorithm on collected data from previous customer activities. In turn, this minimizes prediction errors and enables better-personalized suggestions of computer system configuration. We took advantage of the implicit feedback approach while modeling customer behaviors if they liked or disliked a particular product. After customer behavior data is collected, we form the customer-product matrix and generate personalized product suggestions for each customer with the help of user-item-based collaborating filtering and item-item-based collaborating filtering algorithms. Customer-website interaction is considered a key input variable in creating personalized recommendations. Customers are supposed to use the website and leave interaction data regarding product configurations they’re interested in. To prove the efficiency of this methodology, we developed a prototype application. The product suggestion success rate of the application is tested on datasets generated from log data of the Casper website. Performance results prove that the developed methodology is successful.

Answering questions, finding the most appropriate answer to the question given by the user as input are among the important tasks of natural language processing. Many studies have been done on question answering and datasets, methods have been published. The aim of this article is to reveal the studies done in question answering and to identify the missing research topics. In this literature review, it is tried to determine the datasets, methods and frameworks used for question answering between 2000 and 2022. From the articles published between these years, 91 papers are selected based on inclusion and exclusion criteria. This systematic literature review consists of research analyzes such as research questions, search strategy, inclusion and exclusion criteria, data extraction. We see that the selected final study focuses on four topics. These are Natural Language Processing, Information Retrieval, Knowledge Base, Hybrid Based.

Studying manufacturing production process via data-driven approaches needs the collection of all possible parameters that control and influence the quality of the final product. The recorded features usually come from different steps of the manufacturing process. In many cases, recorded data contains a high number of features and is collected from several stages in the production process, which makes the prediction of product quality more difficult. The paper presents a new data-driven approach to deal with such kind of issues. The proposed approach helps not only in predicting the quality, but also in finding to which stage of the production process the quality is most related. The paper proposes a challenging case study from ArcelorMittal steel industry in Luxembourg.

An agent has to form a team at run-time in a dynamic environment for some tasks that are not completely specified. For instance, the mass of an object may not be given. Estimating the mass of an object helps in determining the team size, which can then be used for team formation. It has recently been shown that the mass of an object can be estimated from its image. In this paper we augment the existing image2mass model with material embedding. The resulting model has been extensively tested. The experimental results indicate that our model has achieved some improvements on the existing state-of-the-art model for some performance metrics.

Artificial Intelligence for digital REStoration of Cultural Heritage (AIRES-CH) aims at building a web-based app for the digital restoration of pictorial artworks through Computer Vision technologies applied to physical imaging raw data.In previous work [7], it was shown that it is possible to develop a multidimensional deep neural network capable of inferring the RGB image from an X-Ray Fluorescence raw data.The developed network comprises two branches: a one-dimensional branch, which works pixel-by-pixel, and a two-dimensional branch, capable of performing image segmentation.In this project, we report the results of the hyperparameter optimisation of both branches.

In this paper we introduce a computational procedure for measuring the semantic readability of a segmented text. The procedure mainly consists of three steps. First, natural language processing tools and unsupervised machine learning techniques are adopted in order to obtain a vectorized numerical representation for any section or segment of the inputted text. Hence, similar or semantically related text segments are modeled by nearby points in a vector space, then the shortest and longest Hamiltonian paths passing through them are computed. Lastly, the lengths of these paths and that of the original ordering on the segments are combined into an arithmetic expression in order to derive an index, which may be used to gauge the semantic difficulty that a reader is supposed to experience when reading the text. A preliminary experimental study is conducted on seven classic narrative texts written in English, which were obtained from the well-known Gutenberg project. The experimental results appear to be in line with our expectations.

Predictive maintenance solutions have been recently applied in industries for various problems, such as handling the machine status and maintaining the transmission lines. Industrial digital transformation promotes the collection of operational and conditional data generated from different parts of equipment (or power plant) for automatically detecting failures and seeking solutions. Predictive maintenance aims at e.g., minimizing downtime and increasing the whole productivity of manufacturing processes. In this context machine learning techniques have emerged as promising approaches, however it is challenging to select proper methods when data contain imbalanced class labels.In this paper, we propose a pipeline for constructing machine learning models based on Bayesian optimization approach for imbalanced datasets, in order to improve the classification performance of this model in manufacturing and transmission line applications. In this pipeline, the Bayesian optimization solution is used to suggest the best combination of hyperparameters for model variables. We analyze four multi-output models, such as Adaptive Boosting, Gradient Boosting, Random Forest and MultiLayer Perceptron, to design and develop multi-class and binary imbalanced classifiers.We have trained each model on two different imbalanced datasets, i.e., AI4I 2020 and electrical power system transmission lines, aiming at constructing a versatile pipeline able to deal with two tasks: failure type and machine (or electrical) status. In the AI4I 2020 case, Random Forest model has performed better than other models for both tasks. In the electrical power system transmission lines case, the MultiLayer Perceptron model has performed better than the others for the failure type task.

During the COVID-19 outbreak, fake news regarding the disease have spread at an increasing rate. Let’s think, for instance, to face masks wearing related news or various home-made treatments to cure the disease. To contrast this phenomenon, the fact-checking community has intensified its efforts by producing a large number of fact-checking reports. In this work, we focus on empowering knowledge-based approaches for misinformation identification with previous knowledge gathered from existing fact-checking reports. Very few works in literature have exploited the information regarding claims that have been already fact-checked. The main idea that we explore in this work is to exploit the detailed information in the COVID-19 fact check reports in order to create an extended Knowledge Graph. By analysing the graph information about the already checked claims, we can verify newly coming content more effectively. Another gap that we aim to fill is the temporal representation of the facts stored in the knowledge graph. At the best of our knowledge, this is the first attempt to associate the temporal validity to the KG relations. This additional information can be used to further enhance the validation of claims.

Preliminary results of the multi-hazard (seismic, landslide, floods) analysis, conducted on some Critical Infrastructures (CI), are presented. Lifelines an CI such as roads, railways, gas and power networks are essential for the Country’s economy, but also to ensure the overcoming of emergencies in those territories in case of natural disasters. ENEA CIPCast platform, an innovative Decision Support System (DSS), has been used for the GIS-based analysis of physical impacts induced by natural hazards on CI. It has been exploited in several projects, including RAFAEL that, by means of analytical and geoprocessing tools, was focus on the extensive assessment and mapping of CI hazard. Geospatial layers, describing and classifying the different hazards in Italy, has allowed to produce a set of specific hazard maps for each CI considered. Multi-hazard values were assigned by using a suitable matrix having four classes. Preliminary results are discussed focusing on the impact on CI exposed to different level of seismic, flood and landslide hazard.

With the introduction of Metaverse, its use is increasing day by day. In this study, the factors affecting the historical development of Metaverse, the Metaverse architecture and the use of Metaverse in the field of education are discussed. The strengths and weaknesses of the use of Metaverse in the field of education are emphasized; the opportunities it will offer and the problems and threats that may be encountered are examined. Since Metaverse is a new concept, a resource for the use of Metaverse in the field of education has been tried to put forward by using the limited number of sources currently available in the literature. In this study, it is emphasized that the Metaverse environment can add a new dimension to the field of educational technologies. However, it should be taken into account that the necessary technologies and architectures in this field are not mature enough yet. Therefore, it is considered a necessity to determine appropriate strategies for the use of Metaverse in the educational field and to start determining its widespread effect until the infrastructure of Metaverse matures.

The main purpose of the article is to highlight the peculiarities of the formation of the system of continuing education in Ukraine. The evolution of the views of scientists of different periods on the need for continuing education is analyzed. The influence of government institutions on the creation of regulatory and legal and organizational support for the implementation of state policy for the development of continuing education in modern Ukraine. The role of the state in the implementation of primary and secondary school reforms for the formation of quality knowledge as a basis for the development of the individual and his need to learn throughout life. The article analyzes the effect of legal and organizational mechanisms of public management of the system of continuing education in the conditions of innovative development of society and implementation of educational reforms. The main problems of legal and organizational nature regarding the formation of a single educational space as a favorable environment for the development of continuing education are identified. On the example of educational institutions of Ukraine the efficiency of mechanisms of public management of the system of continuous education and quality assurance of education in the conditions of application of innovative technologies of education is estimated.

Decentralized computing and blockchain technology play a significant role in the implementation of modern digital economy business models. The most noticeable trends in this economy are the diversification and convergence of platforms and services, which is often achieved through undesirable fragmentation of the overall IT landscape. Business solutions represented by different blockchain networks turn out to be poorly connected, data exchange between them is difficult. The search for ways to overcome barriers between different decentralized networks leads to an increase in the importance of cross-platform integration solutions, providing the necessary level of interoperability. Such solutions must be secure both in terms of confidentiality and fault tolerance. Below is a vision of the architecture of integration gateways using the ODAP-2PC protocol, which provides crash fault-tolerance for the integration of various networks. This architecture provides transparency of interaction, reliability and continuity of audit in digital asset exchange systems or payment systems with increased requirements for interoperability.

In this article we review NFT token architecture and metadata storage options. We show the main disadvantages and risks of the existing NFT architecture and infrastructure. NFT minting performance was analysed on various distributed ledger technologies. The main problem of NFT technology that we highlight is its associated metadata storage options. Since in-secure Web2 based servers are allowed as a storage solution data that is associated with the NFT tokens can be corrupted, modified and lost. Thus harming the end user. We see the main solution for such issues in data blockchains proliferation as they would provide data immutability.

Decentralized finance (DeFi) is now a massive aggregator of many financial blockchain protocols. The overall value locked in them is currently valued at around 80 billion USD. Every day, a growing number of new users bring their investments to DeFi. Decentralized finance entails the creation of a single ecosystem of many blockchains that interact with one another. To enable DeFi, the problem of combining and interacting blockchains becomes critical. In this paper we describe the concept of a DeFi protocol, which employs cross-blockchain interface technologies, and show how a prototype pre-sale application based on the proposed concept can be created.

The article presents the issue of describing complex composite objects of intellectual activity (OIA), which are placed and maintained in information systems using distributed ledger technology (blockchain). The essence of fragmentation is described, a classification of the stages of maturity of deposited OIAs and their structural configurations is introduced. An approach to decomposition and internal markup of composite objects is described. Examples of depositing such objects on the IPUniversity platform (a composite OIA markup module) are given, as well as recommendations on organizing the work of registries operating with complex objects of intellectual activity.

Metaheuristic algorithms have been used for different optimization problems and many modifications and hybridizations of these algorithms have been proposed. One such algorithm, Particle Swarm Optimization (PSO), has been proposed and modified for many distinct problems. Solving systems of nonlinear equations is one of its many applications, but as these systems grow, the effectiveness of PSO and PSO-based algorithms decrease. As such, there need to be modifications that impact the performance of the algorithm, such as increasing the number of particles or the number of iterations. However, there are problems where the combined use of both of these strategies does not solve all the drawbacks associated with the use of these algorithms, so a possibility would be to reduce the search space of the problems considered. In this article, the effect of the search space segmentation for solving nonlinear systems of equations using PSO is explored, and an experimental comparison is made between a simple segmentation of the search space to an increase of the number of particles.

Chillers are the basis of modern refrigeration systems of large facilities, such as oil refineries, power plants and large commercial buildings. The increasing concerns about the scarcity of water and energy resources require careful optimization processes to achieve energy efficiency in industrial buildings. Optimization require mathematical models of real equipment. In this paper, we present two models for a compression chillers, which are one the main equipment in industrial refrigeration systems. We prove that proposed models are precise and faithful to the real compression chiller used in modern refrigeration system. Moreover, we prove that the model’s values model accurately the actual values of the global requirements in terms of power consumption of the whole refrigeration system composed of cooling towers, fans and chillers. The models of the cooling tower and corresponding fans are presented in [5].

Cooling towers and chillers are the basis of modern refrigeration systems of large facilities, such as oil refineries, power plants and large commercial buildings. The increasing concerns about the scarcity of water and energy resources require careful optimization processes to achieve energy efficiency in industrial buildings. Energy efficiency oriented optimizations require mathematical models of real equipment that compose the refrigeration systems. In this paper, we present a complete model cooling towers and corresponding fans based on Merkel’s and Braun’s methods. We prove that proposed model is accurate and faithful to the real cooling cells in modern refrigeration systems. The obtained value of mean square error when applying the model and comparing the obtained results to the actual ones is minimal, hence validated the proposed model.

One of the keys to understanding and treating Alzheimer’s disease (AD) and dementia is believed to be calcium (Ca $$^{2+}$$ 2 + ) in the cytoplasm. Researchers have discovered how an imbalance of Ca $$^{2+}$$ 2 + ions in the cytoplasm can lead to cell death and, more particularly, neurodegeneration in brain cells in Alzheimer’s and dementia patients. Many substances are present in brain cells, but Ca $$^{2+}$$ 2 + is the most tractable and is employed for experimental validations. In this study, we employ a spatio-temporal computational model to investigate AD development using Ca $$^{2+}$$ 2 + dynamics in the cytoplasm. We study the spatio-temporal dynamics of biochemical processes via a new coupled model and analyze the sensitivity of this model to some of the critical parameters. As a result of this study, several important contributions have been made. Firstly, we have demonstrated that the SERCA pump flux parameter has a significant impact on the frequency of intracellular calcium concentrations. Furthermore, we studied Ca $$^{2+}$$ 2 + dynamics with diffusion in the presence of different amyloid-beta levels. We found that how amyloid-beta affects various fluxes contributions through voltage-gated calcium channels, amyloid-beta-formed channels and ryanodine receptors. This work contributes to a better understanding of the spatio-temporal action of Ca $$^{2+}$$ 2 + dysregulation in the cytoplasm, and through this, it can offer further insight into AD developments and progression.

The recent years were rich in new and unexpected social and political factors for Poland, such as the COVID-19 lockdown in 2020–2021 and the refugee crisis in 2021–2022. These ‘wildcards’ will definitely have serious consequences for people and cities, directly and through the impact of so-called externalities. The paper identifies trends in the geographical development of urban areas in Poland during the last five years (2016–2021), particularly in terms of residential suburbanization and urban sprawl. The study aims to explore the driver factors that determine the spatial scale of suburbanization and reveal ‘wildcards’ that may indirectly affect this process but are hard to be quantified and embedded into spatial analysis. Both wildcards and externalities of suburbanization seem to be underexplored, and this paper’s goal is to bring progress on this pass. The spatial analysis applying location quotients (LQ) metrics creates the possibility for comparisons of locations with intensified urbanization for different time moments, thus fulfilling a function similar to the standardization of features considering time and space perspectives. The results makes the evidence to progressive suburbanization around the main Polish cities during the years 2016–2021, revealing, at the same time, distinguishing features of spatial development for the period associated with social and political stresses (2021).

The new mobility paradigm, the spread of new city models and the recent pandemic have highlighted the variability and complexity of people’s mobility. There is a widespread need for dynamic urban planning strategies to adapt to circumstances and mitigate the risk of dangerous events such as natural disasters and pandemics. The last decade has been characterised by urban transformation processes that have often led to inequalities in spatial and infrastructural distribution and services. To reduce this criticality, it is necessary to investigate the reciprocal influence between the spatial configuration of the built environment and the way everyday mobility is implemented and becomes habitual, and its relationship with the inequalities mentioned above (e.g. suburbs, degraded areas, urban voids). This research focuses on studying the evolution of the accessibility of the small mountainous urban centre of Zafferana Etnea, which seismic phenomena have hit several times. In the first instance, the work identified and characterised the case study by estimating the mobility demand based on access and the analysis of the spatial configuration (buildings, infrastructures and services). Similarly, the research proposes the definition of the primary attractive nodes of the city, zoning and the purpose of key factors for the definition of mobility indicators that allow the realisation of a quantitative analysis of transport demand and supply based on indicators rather than on the configuration of the urban fabric, which will be addressed with theoretical and methodological tools provided by the spatial syntax.

Urban security plays a fundamental role in achieving sustainable cities. The environmental approach to security proved crucial, with its inner multidisciplinarity, in fighting crime and in promoting crime prevention through environmental design, urban management and situational-oriented strategies, where the importance of surveillance emerges. Among the research fields involved, digital technology has a primary role in the research progress both through the development of advanced analysis and decision support systems (DSS), and through the elaboration of high-tech interconnected devices in a smart-safe city logic. In this vein, the paper first reviews the IT and ICT role in urban security progress, and then it traces the development and the importance of surveillance by focusing on the isovist and the space syntax theory until their last findings. Finally, the paper suggests a schematic proposal of a digital surveillance-city model based on the integration of those concepts, which could be used for analytic, decision support and predictive purposes. The strength of this model resides in its adaptability to different contexts and its potential independence from sensitive data, which permits overcoming privacy issues. Moreover – being surveillance and visibility features that impact different urban life aspects – this model could represent an urban planning DSS for strategies that goes beyond the mere crime-fighting. Furthermore, this model constitutes a first step in the development of a surveillance-oriented Agent-Based Model.

The space matters: beyond a mere background for economic activities placement, it constitutes a crucial element for their overall dynamism. This rationale, however, contravenes more traditional urban and regional economics approaches that interpret space as intangible within their spatial models. In that vein, notable constraints can be found in economic-based methods and spatial units oriented to spatialize the territorial endowments and interpret their role within the urban structure. While economics’ methods are limited in their spatial representation, urban and regional planning has otherwise focused in providing instruments that address spatial characteristics of areas where urbanization is predominant. Methods that highlight the configurational properties and the organization of the cityscape structure, could aid economics in its methodological issues. Still, a divide persists between these two fields of research, as neither consistently incorporates the methods and variables considered by the other. In this paper we showcase a method used to create a spatial unit oriented to territorialize economic-based datasets represented at a regional scale within the confines of urban areas’ industrial agglomerations. In this proof of concept, variables related to labour – average number of firms, employees, and firm-size; and to installed capital – average real estate prices – are spatialized to outline their patterns distribution across Tuscany’s cities. Organized in a GIS-based environment, the representation of such variables within a computerized space and with a proper spatial unit provides a basis that can be associated to the configurational aspects of the territory, being a complementary analysis for urban and regional planning.

This paper focuses on how a city’s configurational patterns impact the city-wide and neighbourhood spatial, economic, and cognitive context through the case study of Vienna. The authors investigate Vienna by applying the space syntax method to get a better grasp of the city-wide and local to-movement, through-movement potentials, and intelligibility. This approach allows the authors to determine the degree of street life and liveliness of Vienna in relation to walkability potential, which includes social and economic factors. The case study of Vienna is performed using quantitative analyses, with a mathematical street network modelling approach and statistical analyses. Additionally, this quantitative approach is enriched with a qualitative photographic survey. The data imply that Vienna, as a historically developed city, has a high potential for walkability. This is also confirmed by the balance between the foreground network for long-distance movement via motorised public transport, trams, and cars and the background network for walkability in neighbourhoods. The paper concludes by juxtaposing socio-spatial potentials with realised walkability and influencing factors that support or hinder walkability, and by considering how a sustainable urban future can be achieved through well-functioning strategic planning guidelines.

In this paper we focus on the study of monic polynomials whose coefficients are quaternions located on the left-hand side of the powers, by addressing three fundamental polynomial problems: factor, evaluate and deflate. An algorithm combining a deflaction procedure with a Weierstrass-like quaternionic method is presented. Several examples illustrate the proposed approach.

The paper is focused on intrinsic properties of a one-parameter family of non-symmetric number triangles $$\mathcal {T}(n),\;n \ge 2,$$ T ( n ) , n ≥ 2 , which arises in the construction of hyperholomorphic Appell polynomials.

Due to its ability to fit skew implied volatility profiles of market option prices, the constant elasticity of variance (CEV) model overcomes a shortcoming of the constant volatility assumption used by the Black-Scholes lognormal model. Although the CEV model has an analytical formula for European options, the presence of the non-central Chi-square distribution in the formula brings instability in the computation of European option prices for certain parameter ranges. For efficient valuation of the option prices for all realistic option parameters, a sixth-order finite difference method is proposed in this work. The computational procedure employs a seven-point discretisation of the space derivatives on non-uniform grids together with a mesh-refinement technique around the option’s strike price. This new technique is very fast and is shown to achieve sixth-order numerical convergence ratios. The stability of the finite difference method is demonstrated through its capability to generate accurate prices for strongly negative elasticity factors. The numerical method is thus a superior technique to existing methods for numerical computations of CEV option prices.

This paper deals with the dynamics of a special two-parameter family of coquaternionic cubic maps. By making use of recent results for the zeros of one-sided coquaternionic polynomials, we analytically determine the fixed points of these maps. Some numerical examples illustrating the theory are also presented. The results obtained show an unexpected richness for the dynamics of cubic coquaternionic maps when compared to the already studied dynamics of quadratic maps.

In this paper we focus on the study of monic quadratic polynomials whose coefficients are coquaternions and present several new results concerning the number and nature of its zeros. Examples specially constructed to illustrate the diversity of cases that can occur are also presented.

In several clinical and epidemiology studies, data from events that occur successively in time in the same individual, are frequently reported. Among these, the most common are recurrent events where each subject may experience a number of failures over the course of follow-up. Examples include repeated hospitalization of patients, recurrences of tumor, recurrent infections, among others. In this work, the interest is to study the correlation between successive recurrent events, gap times, in the presence of right censoring. To measure the association between two gap times we use the Kendall’s $$\tau $$ τ correlation coefficient, by incorporating suitable bivariate estimators of the joint distribution function of the gap times and of the marginal distribution function of the second gap time, into the integrals that define the probability of concordant pairs and the probability of discordant pairs. Two of the estimators of the joint distribution function of the gap times considered in this work are already known, but we consider also estimators with Kaplan-Meier weights defined by using decision trees and random forests methodology. We conclude that all the estimators perform better in a scenario of negative association. When the association is moderately negative, the performance of the estimator with smoothed weights using random forests is superior. In the case of strong positive association, the best estimator is the presmoothed nonparametric but, in the case of moderate positive association, this estimator has identical performance as the estimator with presmoothed weights using random forests.

The bootstrap method is a widely used tool for quantifying the uncertainty associated with a given statistical estimator or machine learning method. This paper proposes a novel approach for sequential experimental design that uses the bootstrap in conjunction with an interpolating surrogate model. Consider the problem of fitting a surrogate to a computationally expensive simulation model that yields a numerical output given the values of a set of continuous input variables. To fit a surrogate, initial data points are obtained by running the simulations at a set of space-filling design points. The proposed Bootstrap-Surrogate method improves on this experimental design by sequentially identifying points where the surrogate prediction uncertainty is high and then evaluating the simulation at those points. The surrogate prediction uncertainty at a candidate simulation point is estimated using a weighted combination of two criteria, one based on the bootstrap standard error of the surrogate predictions at the candidate point and the other based on the minimum distance of the candidate point from previous design points. The method is implemented using a radial basis function (RBF) surrogate and tested on groundwater bioremediation models and several test problems. The results show that the Bootstrap-RBF approach generally yields better experimental design points for surrogate modeling (as measured by RMSE on a large test set) than those obtained by an optimum Latin hypercube sample or a sequential experimental design based on a maximin criterion on the groundwater bioremediation models and on the test problems with similar structures.

Forecasting in time series is one of the main purposes for applying time series models. The choice of the forecasting model depends on data structure and the objectives of the study. This study presents a comparison of Box Jenkins SARIMA and Holt-Winters exponential smoothing approaches to time series forecasting to increase the likelihood of capturing different patterns in the data (in this specific case, home insurance data) and thus improve forecasting performance. These methods are chosen due to their ability to model seasonal fluctuations present in insurance data. The forecasting performance is demonstrated by a case study of home insurance monthly time series: total and frequency rate time series. In order to assess the predictive and forecasting performance of the two methodologies adopted, several evaluation measures are used, namely MSE, RMSE, MAPE, and Theil’s U-statistics. A comparison is made and discussed, and the results obtained demonstrate the superiority of the SARIMA model over the other forecasting approach. Holt-Winters also produces accurate forecasts, so it is considered a viable alternative to SARIMA.

This paper presents a new approach to density-based clustering for the identification of dense areas. In particular, the focus is on identification of breast masses in the X-ray imaging of a mammography. The idea was to apply cluster analysis by identifying breast masses as clusters, understood as dense regions of space separated by areas of lower density. Attention was focused on a particular method of clustering based on density, the DBSCAN, proposing a new approach by applying it to a real dataset: a supervised approach, based on ROC curves and a weighted distance, for the choice of input parameters.

For decades, researchers and health professionals have been using fetal and newborns measurements to evaluate its development. In recent years, there have been new studies suggesting that the placenta’s measurements and its evolutions are capable of reflecting changes in the fetus’s development and even newborn and adult diseases.Most of these analyses are done using growth curves that use linear regression methodologies such as previous studies done. To account for errors associated with this regression and use a more robust method, quantile regression is used to create the placenta’s growth curves. The dataset used for this study was collected on Portuguese CGC Genetics and involves the Portuguese parturient population from different regions.It is also an objective of this study to create a dynamic application that allows the researcher or health professional to enter placental growth values and compare them to the created growth curves to evaluate the evolution of the placenta. This application uses a CSV file with the information gathered from the placenta and is uploaded to the application which then plots the values on the created growth curves. The application also allows the user to edit the values. This application was created on Shiny and can be accessed at https://samuelalves.shinyapps.io/APP2/ .

Delirium is a common manifestation of severe acute neuropsychiatric dysfunction prevalent in hospital settings, which due to the complex multi-factorial causes is often under-diagnosed and neglected. Early detection of delirium is a critical concern that can be effectively addressed using machine learning (ML) techniques. As such, some methods to improve the accuracy of ML classification models for the detection of delirium are covered in this document. The aim of this paper is to develop and validate a tool for use in a hospital setting to accurately identify delirium during the admission of a patient. A database collected at a Portuguese hospital between 2014 and 2016 was used to conduct this experimental research. Available data comprised 511 records and 124 variables, including patient demographics, medications administered, admission category, urgent admission, hospitalization period, history of alcohol abuse and laboratory results.The methodologies used included data pre-processing, data imbalance processing, feature selection, train and test model with different ML classifiers, evaluating model performance and development of a Python web-based application.The model achieved consists of 26 predictors assessed during admission to a healthcare facility. This model combines the SelectFromModel method with the logistic regression algorithm, resulting in an area under the receiver operating characteristic curve of 0.833 and an area under the precision-recall curve of 0.582.Although the prediction model can be enhanced, this approach could be a useful support tool to identify patients at increased risk for delirium in healthcare settings. The application developed is available on: https://bit.ly/3waT3T7 .

Drowsiness at the wheel has been studied for different countries since it is important for road safety and its prevention. Since it is considered a public health problem, solutions must be found to avoid worse scenarios and to identify a low-cost system.Therefore, this work aims to detect the drowsy state, without labeling it manually, considering the heart rate variability. To make this possible, driving simulations were performed, using a wearable device. In terms of methodology, multivariate statistical process control, considering principal component analysis, was implemented, and compared with a similar study. Three principal components were computed taking into consideration time, frequency, and non-linear domain, every two minutes. Thereafter, Hotelling $$T^2$$ T 2 and squared prediction error statistics were estimated. These statistics were estimated considering each principal component, individually. Thereby, the results achieved seemed to be promising to identify drowsiness peaks. However, the study developed has limitations, like the identification of points out-of-control occurred due to signal noise and it does not identify all the drowsiness peaks. Conversely, it was not used information from the participants’ awake states as a reference. Therewith, new simulations must be done, and new information must be added to avoid noise and to detect more drowsiness peaks.

Point clouds represent the as-is geometry of indoor and outdoor environments by sets of 3D points. They allow for constructing 3D models of objects, sites, cities, and landscapes and, hence, form the base data for almost any conscious, smart city system and application. For implementing such systems, we need a spatio-temporal data structure that enables efficient storage and access to 4D point clouds. In particular, the data structure should allow continuous updates, change tracking, and support for spatial and spatio-temporal analysis. This paper discusses challenges and approaches for a 4D point cloud data structure. In particular, the challenges arise from repeated scanning of environments in terms of sparsity, data redundancy, and geometric blurring of the corresponding point clouds. We outline a scheme for incremental storage of 4D point clouds via signed distance fields using a sparse, voxel-based representation. To efficiently implement analysis operations, we discuss how the data structure supports access based on both spatial and temporal criteria. In particular, we outline how machine learning-based interpretations used to classify point clouds and derive object-based information can work with the data structure.

This paper proposes a global model of sustainable desalination in the hydrological context of any river with a reservoir close to the sea. For this purpose, a desalination plant is installed, located on the coast, which transports the pumped water using the course of the river itself and discharges it directly into the reservoir. The concept of sustainability lies in the fact that the energy required to provide for the entire process is obtained by means of a floating photovoltaic park located in the reservoir. The electrical energy generated by this park provides a self-consumption solution for both the desalination plant and the pumping stations, either by injecting and balancing its energy into the electrical grid or directly to the desalination plant. In addition, as a remarkable product in the desalination process, various derivatives are obtained and the project becomes a source of green hydrogen.

A meta-analysis is the statistical pooling of the summary statistics from several selected studies to estimate the outcome of interest. A job’s salary estimate is important information for both job applicants and companies, that is reported on different websites. By combining data from different sources a mean estimate more representative of the target population can be obtained, especially when the data has high variability and dependence on different factors, as in the salary case. However, data are not reported in each source by the same statistics values. Sometimes, the data are summarized by reporting the sample median and one or both of (i) the minimum and maximum values and (ii) the first and third quartiles. Additionally, the sample size is not always reported. In this paper, we aim to provide a step-by-step process to estimate the salary mean and its confidence interval by combining the data from different sources. We illustrate the process with an example dataset of seven job titles. The performance of two different alternatives to estimate the sample mean is evaluated, and the variation in the outcomes between websites is discussed. A high range of other quantitative data could benefit from the proposed process to obtain an estimate representative of the target population.

In order to numerically simulate the contact between equipment and the skin, it is necessary to have equations that satisfactorily reproduce the mechanical behavior of the skin. Taking into account the variability of the mechanical behavior of the skin, it is desirable that the determination of the coefficients of these same equations and the limits of load application are based on experimental results. The mechanical behavior of the skin depends on the place where the contact occurs, the gender and age of the individual, among other factors, as well as the test parameters used, which makes it difficult to obtain these coefficients and limits. The objective of this study is to verify if it is possible to reproduce the mechanical behavior of the skin by performing the ANOVA approach of the experimental data and verify if the limits of load application in safety and comfort remain stable for a group of individuals. Indentation test that allows to obtain the force vs deformation curve in which the maximum reached is when the maximum pain is reached and where the instant where the pain threshold occurs (pain onset) is also recorded. The test was performed at a specific point on the forearm, with a spherical tip indenter, with a diameter of 5 mm, at a speed of 1 mm/s in eighty (80) healthy subjects, aged between 20 and 28 years, 40 women and 40 men.

The forest surveillance carried in Portugal presents some difficulties. To improve and solve some of these problems the use of new technologies such as unmanned aerial vehicle systems (UAVS) can be implemented in an efficient way. To do so we need to determine the risk of occurrence of a forest fires at a certain time in a certain region. With such goal, we have used several statistical techniques, such as ARIMA or GLM approach. We built adequate models in certain regions but for others it was impossible to determine a good quality model. The work is still going on, we expect to enlarge the area where we can get a good quality prevision of risk fire.

Nowadays, the number of people using the internet online increases day by day. Therefore, there is a growing need to analyze user behavior trends using navigational clickstream data these days. User behavior can be defined as the collection of the user’s click sequence and transition movements between pages within site. Application areas of user behavior play an important role in many areas, from e-commerce to the banking sector. User behavior is a good argument for analyzing users’ interests. Although there are many research studies in the literature related to user behavior mining, we could not identify a comprehensive literature review on this topic. This paper gives a systematic map of the analysis and mining of web domain users’ activity in this research. We give exact user behavior measures, and publications that we anticipate will help other researchers discover study gaps and possibilities in this paper.

In this work we present a numerical solution of a 3D system of transient and nonlinear partial differential equations. The model arises from larvae, algae and mussel interactions, which takes place in the Pereira Barreto channel, connecting an important Brazilian hydroelectric plant to a river. The mathematical model, proposed in Silva [1] for the two-dimensional case, is composed of three advective-diffusive-reactive equations for species densities coupled with the Navier-Stokes equations to simulate the velocity field of the channel. A numerical discretization of the model is proposed within the framework of the finite element method. Since the problem is advection-dominated, we resort to the well-known stabilization schemes, SUPG and CAU, to obtain oscillations-free solutions. The spatial discretization produces a fully implicit set of nonlinear differential algebraic equations that is integrated numerically in time by the two-step Backward Differentiation Formula of second order, whereas the nonlinear process is solved by a Picard fixed point iteration. The proposed formulation is used to simulate the dynamics of species proliferation and to quantify the golden mussel population in a stretch of the Pereira Barreto channel, with a focus on population control measures. The preliminary results, as well as other considerations related to the problem and the numerical model, are discussed.

Sequential pattern mining algorithms are unsupervised machine learning algorithms that allow finding sequential patterns on data sequences that have been put together based on a particular order. These algorithms are mostly optimized for finding sequential data sequences containing more than one element. Hence, we argue that there is a need for algorithms that are particularly optimized for data sequences that contain only one element. Within the scope of this research, we study the design and development of a novel algorithm that is optimized for data sets containing data sequences with single elements and that can detect sequential patterns with high performance. The time and memory requirements of the proposed algorithm are examined experimentally. The results show that the proposed algorithm has low running times, while it has the same accuracy results as the algorithms in the similar category in the literature. The obtained results are promising.