Computation, Volume 10, Issue 8 (August 2022) – 15 articles

In many small- and medium-sized enterprises (SMEs), defective products are still manually verified in the manufacturing process. Recently, image classification applying deep learning technology has been successful in classifying images of defective and intact products, although there are few cases of utilizing it in practice. SMEs have limited resources; therefore, it is crucial to make careful decisions when applying new methods. We investigated sample size sensitivity to determine the stable performance of deep learning models when applied to the real world. A simple sequential model was constructed, and the dataset was reconstructed into several sizes. For each case, we observed its statistical indicators, such as accuracy, recall, precision, and F1 score, on the same test dataset. Additionally, the loss, accuracy, and AUROC values for the validation dataset were investigated during training. As a result of the conducted research, we were able to confirm that, with 1000 data points or more, the accuracy exceeded 97%. However, more than 5000 cases were required to achieve stability in the model, which had little possibility of overfitting. Full article

Respiratory viruses known as coronaviruses infect people and cause death. The multiple crown-like spikes on the virus’s surface give them the name “corona”. The pandemic has resulted in a global health crisis and it is expected that every year we will have to fight against different COVID-19 variants. In this critical situation, the existence of COVID-19 vaccinations provides hope for mankind. Despite severe vaccination campaigns and recommendations from health experts and the government, people have perceptions regarding vaccination risks and share their views and experiences on social media platforms. Social attitudes to these types of vaccinations are influenced by their positive and negative effects. The analysis of such opinions can help to determine social trends and formulate policies to increase vaccination acceptance. This study presents a methodology for sentiment analysis of the global perceptions and perspectives related to COVID-19 vaccinations. The research is performed on five vaccinations that include Sinopharm, Pfizer, Moderna, AstraZeneca, and Sinovac on the Twitter platform extracted using Twitter crawling. To effectively perform this research, tweets datasets are categorized into three groups, i.e., positive, negative and natural. For sentiment classification, different machine learning classifiers are used such as Random Forest (RF), Naive Bayes (NB), Decision Tree (DT), Logistic Regression (LR), and Support Vector Machine (SVM). It should be noted that the Decision tree classifier achieves the highest classification performance in all datasets as compared to the other machine learning algorithms. For COVID-19 Vaccine Tweets with Sentiment Annotation (CVSA), the highest accuracy obtained is 93.0%, for the AstraZeneca vaccine dataset 90.94%, for the Pfizer vaccine dataset 91.07%, 88.01% accuracy for the Moderna vaccine dataset, for the Sinovac vaccine dataset 92.8% accuracy, and 93.87% accuracy for the Sinopharm vaccine dataset, respectively. The quantitative comparisons demonstrate that the proposed research achieves better accuracy as compared to state-of-the-art research. Full article

This paper discusses the relationship between weather and rice productivity modeled using the Cobb–Douglas production function principle, with the hypothesis that rice production will increase in line with the increase in average rainfall, wind speed, and temperature every month and then decrease if the weather conditions exceed the threshold. As a result, farmers have the risk of losing rice production. To overcome this problem, we try to estimate the value of the risk. The purpose of this study is to estimate the risk of losses that occurred in rice plants due to weather changes. The method used in this study is risk estimation with the Tail Value at Risk (TVaR) approach. In addition to TVaR, it is estimated simultaneously with Value at Risk (VaR) and Conditional Value at Risk (CVaR). This study uses weather data consisting of rainfall data, wind speed, and air temperature collected from geophysical and meteorological data. Meanwhile, yield data were obtained and processed from the Central Statistics Agency and the West Java Agricultural Service. The data used are data from 2008 to 2021. There are two main parts of the results in this study, namely mathematical analysis and data analysis. The mathematical analysis is a risk model derivation process, which includes TVaR risk measurement. The data analysis process is a simulation of the estimated risk of rice production loss. The results obtained from this study are the value of opportunity risk of loss based on the VaR, CVaR, and TVaR approaches. The conclusion of this study is that the rice plants have a risk of loss in the form of reduced yields caused by weather changes. Farmers can plan to overcome this loss problem, by setting up a reserve fund. Risk of loss can be managed through the rice agricultural insurance program. This is in line with the Indonesian government’s program through the ministry of agriculture. Thus, farmers, insurance companies, and the government can manage the risk of losing rice yields. Full article

In this work, active pharmaceutical ingredients release from aerogel particles and active pharmaceutical ingredients transportation processes were investigated. Experimental studies were carried out on the release of various types of active pharmaceutical ingredients from various types of aerogel particles. Release curves were obtained. A hybrid model using the lattice Boltzmann method and a cellular automata approach to simulate the release of active pharmaceutical ingredients from aerogel particles and active pharmaceutical ingredients transport processes is proposed. The proposed model can be used in new drug development, which allows partially replacing full-scale experiments with computational ones, therefore reducing the experimental studies cost. Full article

Detection of minor leaks in oil or gas pipelines is a critical and persistent problem in the oil and gas industry. Many organisations have long relied on fixed hardware or manual assessments to monitor leaks. With rapid industrialisation and technological advancements, innovative engineering technologies that are cost-effective, faster, and easier to implement are essential. Herein, machine learning-based anomaly detection models are proposed to solve the problem of oil and gas pipeline leakage. Five machine learning algorithms, namely, random forest, support vector machine, k-nearest neighbour, gradient boosting, and decision tree, were used to develop detection models for pipeline leaks. The support vector machine algorithm, with an accuracy of 97.4%, overperformed the other algorithms in detecting pipeline leakage and thus proved its efficiency as an accurate model for detecting leakage in oil and gas pipelines. Full article

Tourism forecasting has garnered considerable interest. However, integrating tourism forecasting with volatility is significantly less typical. This study investigates the performance of both the single models and their combinations for forecasting the volatility of tourism demand. The seasonal autoregressive integrated moving average (SARIMA) model is used to construct the mean equation, and three single models, namely the generalized autoregressive conditional heteroscedasticity (GARCH) family models, the error-trend-seasonal exponential smoothing (ETS-ES) model, and the innovative smooth transition exponential smoothing (STES) model, are employed to estimate the volatility of monthly tourist arrivals into Malaysia. This study also assesses the accuracy of forecasts using simple average (SA), minimum variance (MV), and novel smooth transition (ST). STES performs the best of the single models for forecasting the out-of-sample of tourism demand volatility, followed closely by ETS-ES. In contrast, the ST combining method surpasses SA and MV. Interestingly, forecast combining methods do not always outperform the best single model, but they consistently outperform the worst single model. The MCS and DM tests confirm the aforementioned findings. This article merits consideration for future forecasting research on tourism demand volatility. Full article

The article analyzes the possibility and rationality of using proctoring technology in remote monitoring of the progress of university students as a tool for identifying a student. Proctoring technology includes face recognition technology. Face recognition belongs to the field of artificial intelligence and biometric recognition. It is a very successful application of image analysis and understanding. To implement the task of determining a person’s face in a video stream, the Python programming language was used with the OpenCV code. Mathematical models of face recognition are also described. These mathematical models are processed during data generation, face analysis and image classification. We considered methods that allow the processes of data generation, image analysis and image classification. We have presented algorithms for solving computer vision problems. We placed 400 photographs of 40 students on the base. The photographs were taken at different angles and used different lighting conditions; there were also interferences such as the presence of a beard, mustache, glasses, hats, etc. When analyzing certain cases of errors, it can be concluded that accuracy decreases primarily due to images with noise and poor lighting quality. Full article

Recent research uses an index to measure economic resilience, but the index is inadequate because it is impossible to determine which disturbance factors have the greatest impact on the economic resilience of cities. This study aims to develop a new methodology to measure the economic resilience of a city by simultaneously examining unwanted conditions and disturbance factors. The ratio of regional original income to the number of poor people is known as Z and is identified as a measure of economic resilience in Indonesia. Resilience is measured by Z’s position in relation to the unwanted area following a specific level of disturbance. If Z is in the unwanted condition, the city’s per capita income will decrease, and the city will be considered economically not resilient. The results of the analysis show that six levels of economic resilience have been successfully distinguished based on research on 514 cities in Indonesia involving nine indicators of disturbance and one variable of economic resilience during the five-year observation period, 2015–2019. Only 3.11 percent of cities have economic resilience level 1, while 69.18 percent have level 0. Economically resilient cities consist of 4.24 percent of cities at level 2, as much as 3.39 percent at level 3, as much as 3.39 percent at level 4, and as much as 16.69 percent at level 5. The novelty of this research is to provide a new methodology for measuring the economic resilience of cities by integrating unwanted conditions as necessary conditions and disturbance factors as sufficient conditions. The measurement of a city’s economic resilience is critical to help the city government assess the security of the city so the government can take preventive actions to avoid the cities falling into unwanted conditions. Full article

The paper proposes an approach to the classification of lifecycle support automation systems for engineering objects, with the proposed structure of the description of the designed object, using a triple description approach: functional, mathematical, and physical. Following this approach, an algorithm for drawing up a functional description of the lifecycle is described, which is based on the principle of unity of analysis and synthesis of the created system in the design process. The proposed solutions are considered using the traditional aircraft shaping methodology with the application of the airplane make-up algorithm as an example. Furthermore, the architecture of a multiagent platform for structural–parametric synthesis of the object was presented; for convenient usage of this architecture, it was proposed to use classification of design tasks in the form of a design cube. The proposed approach allows obtaining an accurate description of the designed object and the subtasks needed to create it, which can reduce the time of the project. Unfortunately, not all decisions can be automated at the given stage of technical development, but what is possible to automate is enough to achieve a reduction in terms of realization and an acceleration of the prototyping process, as shown in the considered example. The actual reduction throughout the lifecycle of the product ranges from 10% to 21% of the planned time. Full article

Biometrics is the automated identification of a person based on distinctive characteristics, such as fingerprints, face, voice, or the sound of footsteps. This last characteristic has significant challenges considering the background noise present in any real-life application, where microphones would record footsteps sounds and different types of noise. For this reason, it is crucial to consider not only the capacity of classification algorithms for recognizing a person using foostetps sounds, but also at least one stage of denoising algorithms that can reduce the background sounds before the classification. In this paper we study the possibilities of a two-stage approach for this problem: a denoising stage followed by a classification process. The work focuses on discovering the proper strategy for applying combinations of both stages for specific noise types and levels. Results vary according to the type and level of noise, e.g., for White noise at signal-to-noise ratio level, accuracy can increase from 0.96 to 1.00 by applying deep learning based-filters, but the same option does not benefit the cases of signals with low level natural noises, where Wiener filtering can increase accuracy from 0.6 to 0.77 at the highest level of noise. The results represent a baseline for developing real-life implementations of footstep biometrics. Full article

In addition to a wide range of socio-economic impacts, traffic congestion during the era of the COVID-19 pandemic has been identified as a critical issue to be addressed. In urban neighborhoods, the timespan of traffic congestion hazard (H_TC) after the curfew lift is subjected to the commuters’ decisions about home-to-shopping center departures. The decision for departing early or late for shopping depends on both the internal (commuter related) and external (shopping center related) factors. The present study developed a practical methodology to assess the H_TC period after the curfew timings. An online questionnaire survey was conducted to appraise the commuters’ perception about departure time and to assess the impact of eight internal (family size, involvement in other activities, nature of job, education level, age, number of vehicles, number of children, and availability of personal driver) and three external (availability of shopping center of choice in near vicinity, distance to shopping center, and size of the city) factors on their decision. With an acceptable 20% response rate, Chi-square and Cramer’s V tests ascertained family size and involvement in other activities as the most significant internal factors and availability of shopping center of choice as the primary external factor. Age, number of children, and size of the city influenced to some extent the commuters’ decisions about early or delayed departure. Large associations were found for most of the factors, except education level and availability of drivers in a household. Fuzzy synthetic evaluation (FSE) first segregated the commuters’ responses over a four level-rating system: no delay (0), short delay (1), moderate delay (3), and long delay (5). Subsequently, the hierarchical bottom-up aggregation effectively determined the period of highest traffic congestion. Logical study findings revealed that most (about 65%) of the commuters depart for shopping within 15 min after the curfew lift, so H_TC in the early part (the first one hour) of the no curfew period needs attention. The traffic regulatory agencies can use the proposed approach with basic socio-demographic data of an urban neighborhood’s residents to identify the H_TC period and implement effective traffic management strategies accordingly. Full article

Theoretical aspects of methods for calibrating Stokes polarimeters are considered. The prospects and opportunities for implementing the presented methods for calibrating portable polarization systems used in biology and medicine are determined. Based on a comparative analysis, a method for calibrating a portable Stokes polarimeter for medical applications is proposed. The chosen method provides the smallest error in measuring the parameters of the Stokes vector for calculating the parameters of optical anisotropy and researching the polarization properties of biological tissues. A series of experimental research and statistical analysis of the spatial distributions of the polarization parameters of the calibration sample was carried out to use the results for forming the instrument matrix of the developed Stokes polarimeter during calibration. Full article

This paper presents an experimental study that analyzes learners’ visual behaviour and brain activity in clinical reasoning. An acquisition protocol was defined to record eye tracking and EEG data from 15 participants as they interact with a computer-based learning environment called Amnesia, a medical simulation system that assesses the analytical skills of novice medicine students while they solve patient cases. We use gaze data to assess learners’ visual focus and present our methodology to track learners’ reasoning process through scanpath pattern analysis. We also describe our methodology for examining learners’ cognitive states using mental engagement and workload neural indexes. Finally, we discuss the relationship between gaze path information and EEG and how our analyses can lead to new forms of clinical diagnostic reasoning assessment. Full article

Several approaches have been proposed to reduce the gap between software engineering education and the needs and practices of the software industry. Many of them aim to promote a more active learning attitude in students and provide them with more realistic experiences, thus recreating industry software development environments and collaborative development and, in some cases, with the involvement of companies mainly acting as potential customers. Since many degree courses typically offer separate subjects to teach requirements engineering, analysis and design, coding, or validation, the integration of all these phases normally necessitates experience in a project context and is usually carried out in a final year project. The approach described in this article benefits from the close involvement of a software house company which goes beyond the common involvement of a potential customer. Students are integrated into distributed teams comprising students, teachers and IT professionals. Teams follow the agile Scrum methodology and use the OutSystems low-code development platform providing students with the experience of an almost real scenario. The results show that this approach complements the knowledge and practice acquired in course subjects, develops the students’ technical and non-technical skills, such as commitment, teamwork, and communication, and initiates them in the methodologies and development strategies used in these companies. The feedback from the teachers involved, software companies and students was very positive. Full article

The Optimal Reactive Power Dispatch (ORPD) problem consists of finding the optimal settings of reactive power resources within a network, usually with the aim of minimizing active power losses. The ORPD is a nonlinear and nonconvex optimization problem that involves both discrete and continuous variables; the former include transformer tap positions and settings of reactor banks, while the latter include voltage magnitude settings in generation buses. In this paper, the ORPD problem is modeled as a mixed integer nonlinear programming problem and solved through two different metaheuristic techniques, namely the Mean Variance Mapping Optimization and the genetic algorithm. As a novelty, the solution of the ORPD problem is implemented through a Python-DIgSILENT interface that combines the strengths of both software. Several tests were performed on the IEEE 6-, 14-, and 39-bus test systems evidencing the applicability of the proposed approach. The results were contrasted with those previously reported in the specialized literature, matching, and in some cases improving, the reported solutions with lower computational times. Full article