Blekinge Institute of Technology
Not a member yet
13576 research outputs found
Sort by
Robust Federated Learning for Zero-Day IoT Malware Detection via Adversarial and Generative Augmentation
No full textBackground. The rapid growth of the Internet of Things (IoT) has created new opportunities for automation and large-scale data use, but also exposes networks to zero-day malware that traditional, signature-based defenses fail to detect. Federated learning (FL) offers a privacy-preserving alternative to centralized intrusion detection, yet faces practical challenges from non-IID client data, adversarial inputs, and poisoned model updates. Objectives. This thesis explores whether a three-phase FL pipeline can improve zero-day detection in IoT networks by combining adversarial training, generative augmentation, and robust aggregation. A centralized baseline is used as an empirical upper bound for comparison. Methods. The approach uses a Long Short Term Memory (LSTM) autoencoder trained in three phases: benign pre-training, adversarial hardening with First Gradient Sign Method (FGSM) and Projected Gradient Descent(PGD), and conditional Generative Adversarial Network (cGAN) augmentation filtered with KolmogorovSmirnov (KS) tests. At the server, three aggregation strategies are compared: Federated Averaging (FedAvg), Multi-Krum, and Robust Knowledge Distillation (RKD). Evaluation is performed on CICIoT2023 dataset with phase-wise threshold calibration and a fixed metric suite. Results. Adversarial training improved resilience to seen perturbations but transferred poorly to unseen families. KS-gated augmentation helped bring down false positives without reducing recall, although the gains in recall were fairly small. The choice of aggregation strategy also shaped stability. FedAvg adapted quickly but showed volatility, Multi-Krum was robust but conservative, and RKD gave the most steady performance across phases. The centralized baseline gave the highest accuracy, but federated training was faster and preserved privacy. Conclusions. Layered defenses improve resilience in federated anomaly detection, though non-IID client data remains the central obstacle. The findings confirm that FL can enhance IoT security by balancing robustness, stability, and efficiency, while also highlighting directions for future research
Using ChatGPT as a Combined Invoice OCR and Key-Value Extractor
No full textThis paper provides details and analysis of findings from three experiments on the capabilities of the OpenAI chatbot ChatGPT-4 with vision capabilities as a dual-purpose tool for Optical Character Recognition (OCR) and key-value data extraction tasks. Our results could be relevant to any task where one is interested in extracting key information from images with approximately equal complexity, and are scalable to be used for large datasets. We did the main experiments in the OpenAI user interface for the model, and a smaller experiment using the API. The experiment used a dataset comprising 1000 digital invoices alongside 1000 photographic images of real receipts, collected from a broad spectrum of market sectors within Sweden. The main experiment gave us a significant accuracy rate, achieving in extracting critical financial information from the digital invoices. Similarly, when applied to the photographic images of receipts, it maintained a high accuracy level of . The smaller experiment gave us an accuracy of These findings are particularly noteworthy not only because of the high accuracy levels but also due to the model's effectiveness in performing both OCR and key-value extraction tasks as a one-step process. This dual functionality underscores the model's potential as a highly efficient and reliable solution in automating financial data extraction and processing tasks.
Using Dynamic Programming and Reinforcement Learning for Exploring Tradespaces in Changeability Assessment
No full textThe construction machinery industry faces many uncertainties stemming from environmental, operational, and market-related factors. To mitigate future risks, development teams often favor more broadly applicable solutions compared to localized performance gains in specific scenarios. This situation highlights the necessity of incorporating changeability in these solutions for developing value-robust systems that can manage future uncertainty. Changeability assessment relies on an effective tradespace exploration that provides a unified view of different system configurations and control policies. To support the design teams in exploring such tradespaces, this paper presents an approach combining Dynamic Programming (DP) and Reinforcement Learning (RL) for evaluating optimal control policies, illustrated through a wheel loader application. The underlying basis is that the overall task can be decomposed into several sub-tasks to be solved by DP or RL selectively. A control policy combined from these sub-tasks is presented along with an illustrative tradespace mapping system attributes. The results show that by combining the strengths of DP and RL, the proposed approach can be beneficial when exploring a wide range of solutions. It allows direct comparisons between configuration and control policy changes, which is crucial for effective changeability assessment. However, several limitations have been acknowledged and will be addressed in future studies
Examining sleep health and its associations with technology use among older adults in Sweden : insights from a population-based study
No full textIntroduction: Exploring the association between technology use and sleep health in older adults is important as digital engagement becomes integrated into society. Objective: This study aimed to examine sleep health and its association with technology use in a population-based cohort of 60 years and older. Methods: This cross-sectional, population-based study (2023) included 436 older adults from the Swedish National Study on Aging and Care, Blekinge (SNAC-B) population. These participants were sent questionnaires about their sleep, internet usage, Digital Social Participation (DSP), Technology Anxiety (TA), Technology Enthusiasm (TE), and use of information and communication technology. We used a multidimensional instrument, SATED, to measure sleep health. In this study, we conducted statistical analyses using the chi2 test, T-test, Pearson correlation, and backward linear and logistic regression. Results: Our study found that older adults (60 years+) have a mean sleep health score of 7.40 (SD = 2.03). TE (,) and DSP (,) were positively associated with better sleep health, while TA (,) was negatively associated. Frequent internet users(M = 7.6) and engaging with screens before bedtime (M = 7.7) had higher sleep health scores compared to non-frequent users (M = 6.90,) and none or seldom engagement with screens before bedtime (M = 7.10,) respectively. Linear regression showed TE positively associated (= 0.241,) while TA negatively associated (= -0.220,) with sleep health. DSP was found to be a predictor of better satisfaction (OR: 1.32,), efficiency (OR: 1.16,), and duration of sleep (OR:1.16,). Lower TA predicted better satisfaction (OR: 0.81,), timing (OR: 0.74,), and efficiency (OR:0.78,) of sleep. Older adults who use technology one hour before sleep have better sleep timing (OR: 3.003,), while those who do use mobile phones with a screen during the awake period after sleep onset have poor sleep timing (OR:0.016,). Conclusions: DSP and TE support better sleep health, while TA negatively impacts sleep satisfaction, timing, and efficiency. Encouraging positive digital engagement and minimizing technology-related stress may promote healthier sleep in older adults. SNA
Patient-Controlled Sedation During Gynaecological Procedures—Aspects of Patient Satisfaction, Feasibility and Cost Per Patient : Empirical Research Quantitative
No full textAim: To evaluate the use of patient-controlled sedation with propofol in outpatients with anxiety and fear of gynaecological procedures, considering patient satisfaction, feasibility and costs. Design: This quantitative study used a descriptive retrospective design. Review Methods: This study evaluated patient satisfaction, procedure feasibility and direct costs for PCS with propofol for outpatients with anxiety and fear of gynaecological outpatient procedures. Data including patient age and evaluation of the procedure, as well as procedure-related information describing the type of procedure, duration of the procedure, drugs used, vital signs, interventions and evaluations by staff was extracted from medical records. Data Sources: Data was collected from the medical records of all outpatients who completed a gynaecological procedure using PCS with propofol during the period February 2021 until July 2023 at the GC, Vrinnevi Hospital (Sweden). Results: A total of 128 gynaecological procedures with patient-controlled sedation were performed successfully. Few transient cardiopulmonary events occurred (n = 13), and only one needed use of a jaw thrust to resolve desaturation. The feasibility of the procedure and sedation were assessed to ‘Easy’ (84%, 96%) and no sedation-related cancellations of the procedure were reported. Patients' overall satisfaction was high and they were reported to be ‘Very satisfied’ in 86% of the procedures. The mean cost per patient was 85% lower when the procedure was completed at a gynaecological outpatient clinic instead of a daycare surgical unit. Conclusion: Patient-controlled sedation outpatients with anxiety and fear for gynaecological procedures has been shown to be a well-tolerated method of sedation that gives high patient satisfaction, allows procedures to be completed in a high-quality way and has the potential reduce direct costs. Implications for the Profession and/or Patient Care: Patient-controlled sedation with propofol in outpatients with anxiety and fear of gynaecological procedures may improve patient satisfaction and procedure with the potential to reduce direct costs. Impact: The findings of this study contribute to the growing body of knowledge on patient-controlled sedation (PCS) in gynaecological procedures. By demonstrating that PCS with propofol is feasible, safe and associated with high patient satisfaction, this study highlights its potential to reduce patient anxiety and improve the sense of control during procedures. In addition, PCS may decrease the need for premedication with oral benzodiazepines and reduce referrals to day-surgery units, thereby saving healthcare resources and increasing accessibility for patients. Taken together, the results suggest that PCS can be an important tool in enhancing both the quality and efficiency of gynaecological care. Reporting Method: Reporting followed the STROBE checklist for cross-sectional studies. Patient or Public Contribution: No patient or public contribution.
Measuring and Evaluating Preattentiveness in Video Game Settings : A Study of the Viability and Effectiveness of Preattentive Attributes in 3D Environments
No full textPreattentive processing refers to the human brain’s ability to rapidly detect specific visual features before conscious awareness. This study is split into two parts, with the purpose of the first being to evaluate the effectiveness of preattentive processing in identifying targets with features such as motion, color, and form within controlled 3D video game environments. To this end, participants were shown a series of game scenes, each containing a target object designed with one of these preattentive attributes and were tasked with identifying these objects within a 250-millisecond window, the upper threshold of a defined time frame for what can be considered as preattentive processing. The visual complexity of the scenes was kept consistent, ensuring a uniform look and feel. Results reveal that motion and color are highly effective in guiding attention, with participants achieving perfect accuracy. In contrast, form detection was notably less reliable, with greater variability in accuracy. These findings contribute to understanding how visual complexity and uniformity impact preattentive processing in digital environments, with practical implications for designing visual tasks, game environments, and interfaces that guide user attention more effectively. The second part involved participants playing through a short game where they would walk through the world and interact with objects that used one of the preattentive features to stand out from the background. Following the completion of this game, participants would answer a questionnaire regarding how effective they found the attributes to be at conveying the objective of the game, as well as which attributes they preferred to be guided by in a setting such as this. Results reveal that color is preferred over motion, and going one step further the most preferred type of color target was a glowing blue crate, and the most preferred type of motion target was a crate that would float up and down. These findings contribute to linking the fields of preattentive processing and gaze in video games together
Biosignal Sequence Real-time Prediction for Game Users Based on Features Fusion of Local-Global and Time-Frequency Domain
No full textBiosignal sequence real-time prediction (BSRP) is essential for predicting the future emotional experience of game users. However, BSRP for game users faces challenges, including poor real-time performance and limited feature fusion dimensions. To address these issues, we proposed a method for BSRP based on the features fusion of Local-Global and Time-Frequency domain (LGTF) for game users, which integrates real-time capabilities with multi-dimensional features fusion. Specifically, LGTF meets real-time requirements and achieves the features fusion of Local-Global through multi-channel synchronized adaptive convolution. In addition, LGTF implements the features fusion of inter- and intra-band in the frequency domain and the features fusion of time-frequency domain by incorporating the Self-Attention mechanism and Fourier Transform. Furthermore, we conducted comprehensive validation experiments on LGTF using the public dataset. The results indicate that: 1) In the comparison study, LGTF outperformed other methods, achieving the lowest average MSE and MAE values across different prediction lengths of 0.61 and 0.47, respectively. 2) Ablation studies revealed that the addition of time-frequency domain feature fusion (TF) and local-global feature fusion (LG) both have the positive effect on the prediction performance, reducing the average MSE by 0.11 and 0.09, respectively. 3) Generalization study shows that LGTF exhibits stable performance and generalization across different subjects and shows performance advantages in specific game scenarios. 4) Time performance analysis suggests LGTF has the real-time performance.5) Case study demonstrates that LGTF is practical for predicting game users' future emotions and enhancing their emotional experiences
Efficient Event Aggregation using Eiffel Protocol in CI/CD
No full textBackground: In today’s software development, the systems that manage and automate how new software is built and delivered, known as Continuous Integration and Continuous Delivery (CI/CD) pipelines, are becoming more complex. These systems need to deal with many different events that happen during the software development process. Traditional tools used for organizing and understanding these events, like Ericsson’s Eiffel Intelligence tool, often struggle with complicated scenarios and are not easy to use. This makes it hard for developers to get a clearpicture of the software’s development stages, which is crucial for making timely andinformed decisions. Therefore, there’s a strong need for better tools that can handle these challenges more effectively. Objectives: This thesis aims to introduce an alternative event aggregation solutionfor CI/CD pipelines by addressing the limitations of Eiffel Intelligence. The objectives include: Identifying existing event aggregation methods. Analyzingtheir performance and efficiency within CI/CD processes. Implementing a newaggregation model using Apache Kafka and Apache Flink. Comparing the newmodel against Eiffel Intelligence in terms of efficiency, usability, and complexity. Methods: A comprehensive literature review was conducted to evaluate current event aggregation tools. An experimental setup involving Apache Kafka, and ApacheFlink, was implemented to simulate real-time event aggregation. Data from CI/CD pipelines were processed using this setup, and performance metrics such as code complexity and system scalability were measured and analyzed. Results: The new event aggregation tool we created using Java addressed a criticalissue that the existing Eiffel Intelligence tool could not solve, specifically a complex fourth use case that had previously been left unmanaged. This improvement significantly boosts the tool’s ability to handle challenging scenarios within CI/CD pipelines. Our solution makes rule-writing simpler by using straight forward coding methods instead of complex rules, leading to a cleaner and easier-to-understand codebase. The tool successfully processes event data in real time, making it more user-friendly and adaptable for managing CI/CD operations effectively. Conclusion: This thesis presents an effective alternative to the existing Eiffel Intelligence tool, demonstrating that the integration of Apache Kafka and ApacheFlink can enhance event aggregation processes within CI/CD pipelines. The modular design improves the flexibility, scalability, and usability of event handling. This approach offers valuable insights for organizations seeking to streamline their CI/CD processes, particularly in handling large-scale event data
Kampen mot det osynliga hotet : Patienters upplevelser av psykisk ohälsa vid abdominellt aortaaneurysm, från besked till behandling; en litteraturöversikt
No full textBakgrund: Abdominellt aortaaneurysm (AAA) är ett artärbråck på den stora kroppspulsådern i buken som kan leda till dödliga konsekvenser inom några minuter vid ruptur. För att förebygga och reducera dödligheten används screening, en ultraljudsundersökning av kroppspulsådern i buken. Behandling av AAA inkluderar kirurgiska ingrepp, såsom endovaskulär aortakonstruktion (EVAR) och öppen kirurgi, men även farmakologisk behandling för att bromsa aneurysmets tillväxt. Att drabbas av en allvarlig sjukdom är en omvälvande förändring i livet, där vardagen plötsligt präglas av osäkerhet, oro och en förändrad relation till den egna kroppen. Genom att beskriva patienters upplevelser av psykisk ohälsa i samband med diagnosbeskedet och fram till behandling, kan sjuksköterskor få en bredare kunskap kring hur det är att leva med sjukdomen, vilket i sin tur kan bidra till en mer personcentrerad vård. Syfte: Syftet var att beskriva patienters upplevelser av psykisk ohälsa vid AAA, från besked till behandling. Metod: Litteraturöversikt av kvantitativ och kvalitativ forskning utifrån Fribergs metod. Tio kvalitativa artiklar och en kvantitativ artikel analyserades för att framställa ett resultat. Resultat: Fyra huvudkategorier identifierades: Informationsbrist, ångest, oro & rädsla, begränsningar i vardagen och tillit till sjukvården. Slutsats: Patienter med AAA upplever ofta psykisk ohälsa i form av ångest, oro och rädsla. Dessa känslor uppstår som ett resultat av informationsbrist kring sjukdomen, vilket leder till begränsningar i vardagen, främst på grund av rädslan för ruptur. Dessa patienter har ett stort behov av trygghet och försäkran från sjukvården för att kunna hantera sin sjukdom och för att kunna leva ett normalt liv
Modified Kumaraswamy seasonal autoregressive moving average models with exogenous regressors for double-bounded hydro-environmental data
No full textThis paper proposes the MKSARMAX model for modeling and forecasting time series that can only take on values within a specified range, such as in the interval (0,1). The model is especially good for modeling double-bounded hydro-environmental time series since it accommodates bounded support and asymmetric distribution, making it advantageous compared to the traditional Gaussian-based time series model. The MKSARMAX models the conditional median of a modified Kumaraswamy distributed variable observed over time, by a dynamic structure considering stochastic seasonality and including autoregressive and moving average terms, exogenous regressors, and a link function. The conditional maximum likelihood method is employed to estimate the model parameters. Hypothesis tests and confidence intervals for the parameters of the proposed model are derived using the asymptotic theory of the conditional maximum likelihood estimators. Quantile residuals are defined for diagnostic analysis, and goodness-of-fit tests are subsequently implemented. Synthetic hydro-environmental time series are generated in a Monte Carlo simulation study to assess the finite sample performance of the inferences. Moreover, MKSARMAX outperforms beta SARMA, SARMAX, Holt-Winters, and KARMA models in most accuracy measures analyzed when applied to useful water volume datasets, presenting for the first-step forecast at least 98% lower MAE, RMSE, and MAPE values than competitors in the Caconde UV dataset, and 54% lower MAE, RMSE, and MAPE values than competitors in the Guarapiranga UV dataset. These findings suggest that the MKSARMAX model holds strong potential for water resource management. Its flexibility and accuracy in the early forecasting steps make it particularly valuable for predicting flood and drought periods