Vilnius Tech DSpace Repository
Not a member yet
52562 research outputs found
Sort by
Machine Learning and Statistical Techniques for Outlier Detection in Smart Home Energy Consumption
No full textDue to the continuous increase of smart home culture worldwide, large volumes of energy consumption data gained the attention of data scientists. Smart meters capture the energy consumption readings at a predefined rate and store them as a database. The quality of these databases is highly desired to have accurate analysis and decision-making. But, these readings often have anomalies namely missingness, redundancy, and outliers due to the issues present in meter/data communication networks. Among these, outlier readings indicate an abnormality of the load behavior (e.g.: nonlinearity, unpredicted load switching, system faults, etc.). Hence, it is essential to detect and visualize such anomalies for the necessary treatment. With this motivation, this paper implements various key machine learning and statistical techniques namely autoregressive integrated moving average (ARIMA), autoencoder, density-based spatial clustering of applications with noise (DBSCAN), isolation forest, k-means, hierarchical density-based spatial clustering of applications with noise (HDBSCAN), one-class support vector machine (SVM), local outlier factor (LOF), long short-term memory (LSTM), winsorization, interquartile range (IQR), and Z-score. The results revealed that DBSCAN consistently demonstrated the most accurate performance in detecting outliers in energy data, while, Z-score, IQR, and winsorization provided reasonable outcomes but were limited in handling complex and non-linear data patterns. Autoencoder, Isolation forest, and One-class SVM showed moderate success, but their performance depended on the specific dataset characteristics. Kmeans exhibited mixed results. ARIMA, LOF, LSTM, and HDBSCAN had limited success in outlier detection in the timeseries data. Thus, this analysis finally recommends DBSCAN as the best technique as it consistently outperformed other machine learning and statistical techniques in accurately detecting outliers in smart home energy consumption data.Taip / Ye
Real-time customer communication in e-commerce: improving customer experience, satisfaction and loyalty
No full textIn the era of business digital transformation, real-time customer communication has attracted a lot of attention. The purpose of the research is to investigate how real-time customer communication affects the success of ecommerce businesses, with an emphasis on how prompt and efficient communication can raise customer satisfaction, lower cancellation rates and make companies accomplish a competitive advantage. The research is focused on examining various real-time communication channels including social media, Chatbots, live chat, etc. Additionally, the study also goes over the advantages of using an Omnichannel communication strategy in e-commerce, emphasizing the value of a cohesive communication strategy and the difficulties in integrating various communication channels. The research methodology includes an analysis of existing research on the given topic and a survey distributed among e-commerce business owners/managers. This study intends to contribute to the field by giving new insights into the impact of real-time customer communication on e-commerce growth, which would be beneficial for policymakers, researchers, and industry practitioners.Taip / Yes
16th International Scientific Conference “Transbaltica 2024"
No full text“TRANSBALTICA 2024” received more than 100 contributions from 16 countries around the world. After a careful single-blind peer-review process, which involved reviewers from various parts of the world, 60 papers were accepted. Each paper was evaluated by two reviewers: one member of the scientific committee and one external academic reviewer. During the conference, researchers presented their works, addressing a variety of scientific problems within the research fields of transport engineering, transportation, logistics, and other disciplines, as well as interdisciplinary areas related to transport systems
Unveiling risk patterns through an in-depth analysis of Slovak company defaults (2014–2018)
No full textThis article comprehensively analyses Slovak company defaults between 2014 and 2018, seeking to identify non-financial risk patterns across different sectors, regions, company ages, and legal structures. The study encompasses 168,252 Slovak companies, of which 941 experienced defaults. This paper employs descriptive analysis to outline and summarize pivotal characteristics of the data, thereby understanding the overarching trends in company defaults. This method enables for a more informed approach to develop risk mitigation strategies. The main objective of this article is to unveil the non-financial risk patterns among defaulted companies. The scope of this analysis includes companies operating within Slovakia, offering a comprehensive view of the default dynamics within this geographical and economic context. Our findings point to a gradual decrease in company defaults over the study period. Bratislava region stood out as having the highest number of defaulted companies; however, when looking at the proportion of bankruptcies, Kosice region exhibited the highest percentage of business failures. An interesting age-related pattern emerged from our data, showing a significant concentration of defaults among companies aged between 3 to 9 years. Yet, when defaults are examined as proportions within each age group, companies aged 18 to 22 years demonstrated the highest bankruptcy rates. Industry-wise, the construction sector recorded the highest number of defaults, aligning with the general vulnerability of joint stock companies, which showed a higher likelihood of defaulting at older ages. On the contrary, limited liability companies tended to default more frequently at younger stages of their lifecycle. Notably, the information technology sector emerged as the industry with the lowest default rates, highlighting its relative financial stability compared to other sectors.Taip / YesISlovak Scientific Agency VEGAEU COST Action CA19130 FintechAI in FinanceVEGA-1/0639/2
Monitorinf of permanent grasslands using syntetic Aperture Radar (SAR) coherence and intensity bands composition
No full textStraipsnyje pateiktas tyrimas, kuriam naudojami sintetinės Apertūros radaro (SAR) palydovinių vaizdų duomenys. Taikant koherencijos ir intensyvumo bangų kompozicijas atikta daugiamečių pievų stebėsena. Penkerių metų periodo analizė, remiantis kasmetinėmis SAR vaizdų kompozicijomis, leido įvertinti šio metodo tinkamumą. Gautas rezultatas su išskirtinai aukštais tikslumo rodikliais – tikslumas siekia 95,8 %, daugiamečių pievų identifikavimo su kontroliniais duomenimis atitikimas – 97,1 %, o gauta svertinio tikslumo vidurkio F1 vertė – 96,5 %. Aukšti tikslumo rodikliai patvirtina, kad SAR palydoviniai vaizdai yra itin patikimi duomenys ir gali būti efektyviai naudojami nuosekliai (periodinei) daugiamečių pievų stebėsenai vykdyti.This paper presents a study analysing the performance of Synthetic Aperture Radar (SAR) satellite images, specifically their coherence and intensity bands, for monitoring permanent grasslands. An analysis over a five-year period, based on annual SAR image compositions, allowed the validity of this method to be assessed. The exceptionally high accuracy rates – 95.8% precision, 97.1% recall and 96.5% F1 value – confirm that SAR satellite imagery is a highly reliable tool for consistent monitoring of permanent grasslands. The high accuracy rates confirm that SAR satellite images are highly reliable data for continuous monitoring of permanent grasslands.Taip / Ye
Aplinkai nekenksmingų garsą sugeriančių medžiagų, pagamintų iš žemės ūkio atliekų pluošto, įvertinimas
No full textAgriculture, the world’s largest industry, significantly contributes to the GDP of many developing countries, employing over a billion people and producing 1.3 trillion dollars’ worth of food annually. Despite its economic impact, agriculture generates a substantial 140 billion metric tons of waste globally, necessitating sustainable waste management to reduce CO2 emissions. Natural agricultural waste fibers like coconut fiber, groundnut shell, and sugarcane fiber are explored as eco-friendly alternatives for sound insulation to combat noise pollution. The research investigates their application as sustainable sound-absorbing materials, determining sound absorption coefficients based on the ISO 10534-2 standard. Results indicate coefficients ranging from 0.55 to 0.95 within the 160 Hz to 5000 Hz frequency range. Sugarcane fiber exhibited the most favorable coefficients, reaching 0.95 at 1600 Hz and 0.46 at 800 Hz, followed by coconut fiber with a range of 0.84 at 4000 Hz to 0.57 at 160 Hz. This research highlights the potential of agricultural waste fibers in addressing environmental concerns associated with agricultural waste while providing sustainable solutions for sound absorption.Žemės ūkis, kaip didžiausia pasaulio pramonė, reikšmingai prisideda prie daugelio besivystančių šalių BVP, įdarbinant daugiau nei milijardą žmonių, ir kasmet pagamina maisto už 1,3 trilijono dolerių. Nepaisant jo ekonominio poveikio, žemės ūkis pasaulyje generuoja 140 milijardų tonų atliekų kiekį, reikalaujantį tvaraus atliekų tvarkymo, siekiant sumažinti CO2 emisijas. Natūralūs žemės ūkio atliekų pluoštai, tokie kaip kokoso pluoštas, žemės riešutų kevalai ir cukranendrių pluoštas, tyrinėjami kaip ekologiškos alternatyvos garso izoliacijai kovojant su triukšmo tarša. Tyrime nagrinėjama jų, kaip tvarių garsą sugeriančių medžiagų, taikymo galimybė, nustatant garso sugėrimo koeficientus, remiantis ISO 10534-2. Rezultatai rodo koeficientus, kurie kinta nuo 0,55 iki 0,95, 160–5000 Hz dažnių diapazone. Cukranendrių pluošto bandiniai pasižymėjo aukščiausiomis vertėmis (0,95), esant 1600 Hz dažniui, kokoso pluoštas (0,84), esant 4000 Hz dažniui. Šis tyrimas rodo žemės ūkio atliekų pluoštų potencialą sprendžiant su žemės ūkio atliekomis susijusias aplinkosaugines problemas, kuriant tvarius sprendimus garso sugerčiai.Taip / Ye
Do quality and internationalization interrelate in higher education?
No full textThis article presents a methodical review of the literature focusing on the relationship between quality and internationalization. The authors have formulated a research query for the literature review: How do quality and internationalization interrelate? Upon scrutinizing the research articles that closely align with the topic, it was noted that within the realm of quality and internationalization, there exists an inseparable nexus between globalization and competitiveness. Through an examination of the literature, it was deduced that two distinct perspectives exist concerning internationalization and quality. The PESTEL analysis recognized various classifications of elements, with the principal ones being economic, social, and legal factors, leading to the affirmation that each set of elements in the PESTEL analysis is interconnected with quality assurance.Taip / Yes
Implementing a Support Vector Classifier for Student Risk Assessment in Colegio De Getafe: A Machine Learning Approach
No full textThis study addresses the persistent need for technological advancement in higher education to provide quality higher education and early classification for students who are struggling academically. The study's focus is Colegio de Getafe (CDG), a higher educational institution in Bohol, Philippines. The current process of manually collecting and analyzing data employed by CDG is time-consuming and often results in inaccurate assessments. The study emphasizes the importance of implementing machine learning algorithms for risk assessment systems in identifying at-risk students. The research aims to develop a machine learning-based student risk assessment system that automates data processes and classifies at-risk students. The system enables data collection, enhances data management, and facilitates early identification of students facing academic challenges by implementing the support vector classifier that will identify at-risk students. The study evaluated the machine learning algorithm's performance using the confusion matrix, accuracy, precision, recall, and F1 score. It also assessed the system's quality using the International Organization for Standardization (ISO 9126) software quality standards. Results indicate that the system is highly effective in identifying at-risk students, achieving 100% accuracy, precision, recall, and F1 score in student risk classification. The system complies with international software quality standards, scoring an overall rating of 4.38 out of 5. This research provides a strong solution for early identification and support for at-risk students, aligning with CDG's commitment to providing globally competitive students.Taip / Ye
The relationship between business ethics and law: limits and possibilities
No full textThe moral values that form the company’s identity should remain one of the most important aspects in business, often exceeding legal regulation. At all times, the relationship between business ethics and law has balanced between the line to choose for economic benefits, the interests of the company and shareholders, employees, or consumers, whether the company can behave unethically within the limits of the law. This question has become particularly relevant since the outbreak of the war in Ukraine, which has fundamentally forced various corporations to re-examine their beliefs and values, asking whether a law-abiding business can be righteous in war. The purpose of business ethics is to form standards of morally correct behaviour in business, to promote reliability, respect and accountability in organizations. It is business ethics that help enforce the law by specifying acceptable behaviour that no one can control, but which is necessary to make sense of the letter of the law. This article reviews business ethics violations in peacetime and legal regulatory issues and presents aspects of business ethics violations in wartime that are shaping new business limits and opportunities in the future.Taip / Yes