154092 research outputs found
Sort by
Wedge-Parallel Triangle Counting for GPUs
No full textFor fast processing of increasingly large graphs, triangle counting – a common building block of graph processing algorithms, is often performed on GPUs. However, applying massive parallelism to triangle counting is challenging due to the algorithm’s inherent irregular access patterns and workload imbalance. In this work, we propose WeTriC, a novel wedge-parallel triangle counting algorithm for GPUs, which, using fine(r)-grained parallelism through a lightweight static mapping of wedges to threads, improves load balancing and efficiency. Our theoretical analysis compares different parallelization granularities, while optimizations enhance caching, reduce work-per-intersection, and minimize overhead. Performance experiments indicate that WeTriC yields 5.63× and 4.69× speedup over optimized vertex-parallel and edge-parallel binary search triangle counting algorithms, respectively. Furthermore, we show that WeTriC consistently outperforms the state-of-the-art (i.e., on avg. 2.86× faster than Trust and 2.32× faster than GroupTC).</p
Variable Packet Arrival Rates and Activity Durations in Human Activity Recognition with Wi-Fi Channel State Information
No full textWi-Fi channel state information has gained traction for human activity recognition, localization, and physiology monitoring, due to the positive results found. However, most of these solutions use different sampling rates, input durations, and neural networks, making the data scalability and robustness of future adaptation challenging. To that extent, this paper explores using adaptive pooling layers, namely spatial pyramid pooling, to reduce additional weights and training to handle varying packet arrival rates and activity durations. On a self-collected dataset with 20 participants, it is shown that spatial pyramid pooling achieves accurate human activity recognition with changing sampling durations ranging from 0.1 to 10 and packet arrival rates of. 1 to.100 Hz, with an F1-score.> 0.80 in certain scenarios. These observations are validated on three different datasets for human activity recognition and sign language gestures with different collected transmission rates. The evaluation shows a trade-off in accuracy versus scalability for different packet arrival rates and frame durations, along with a discussion on the possibilities of quickly retraining when changes occur in the context of joint communication and sensing in Wi-Fi channel state information systems.</p
Optical technologies for multiparametric heart failure management beyond the hospital
No full textThe rising prevalence of heart failure (HF) and consequent clinical burden on global health care systems intensified the demand for decentralized, innovative, and cost-effective diagnostic tools. Regular assessment of HF-related biomarkers, such as natriuretic peptides, electrolytes, and renal indicators, along with continuous physiological monitoring, is vital to reduce hospitalization and mortality rates. Patient-centered sensing technologies have the potential to revolutionize HF management by enabling better and more personalized prevention, diagnosis, and therapy. This review highlights recent advances in optical sensing technologies for the non-invasive, point-of-care monitoring of HF beyond clinical settings. A critical analysis is provided of key optical technologies - plasmonics, colorimetry, fluorescence, optical fibers, and photoplethysmography, focusing on their analytical performance and integration into personal use platforms. The purpose of this review is to present a synergistic, multiparametric framework for HF monitoring, demonstrating how diverse optical techniques can be integrated to enhance diagnostic value and support decentralized care. Furthermore, the combination of these methods and technologies with innovative and emerging systems such as microfluidics, dermal tattoos, patches, textiles, and smartphone-based interfaces is discussed in the context of enabling real-time, personalized HF management. Current limitations, technological readiness, and future directions for clinical translation are also addressed, offering insights into how optical analytics can reshape chronic disease monitoring beyond the hospital.</p
Land tenure and land use dynamics in the context of pastoral and non-pastoral land use coexistence in Mvomero district, Tanzania
No full textPastoral and non-pastoral land use coexistence across time and space has implications for the capacity of communities to sustain themselves. This study employed a mixed-methods approach to characterise the trends of land tenure and land use land cover (LULC) in relation to land use coexistence between pastoralists and non-pastoralists. The study was conducted in the Mvomero district of Tanzania, among the districts with high land use conflicts between pastoralists and non-pastoralists. Results indicate that land tenure transformation occurred mainly from village land tenure to state land tenure, followed by state tenure to private tenure. Village tenure was irreversibly converted to state and private tenures. LULC trend showed agriculture and bushland increased at the expense of forest, which decreased consistently from 1994 to 2024. Many respondents indicated that land use coexistence is happening informally, and arrangements are organised mainly by individuals and village leaders. In conclusion, the district's land tenure and land use transformations prioritize biodiversity conservation and agriculture expansion, and overlook pastoral land use. Also, the transformations promote the separation of land uses and undermine land use coexistence. These perpetuate land use conflicts and impede the progress toward zero hunger, a sustainable development goal. A framework that harmonizes the arrangements to accommodate land use coexistence is necessary.</p
Comparative Analysis of Machine Learning Algorithms for Phishing Detection
No full textPhishing attacks represent a significant threat in the realm of cybercrime, characterized by the use of deceptive emails and websites to extract sensitive information from unsuspecting individuals. Traditional detection methods have become less effective against increasingly sophisticated phishing techniques, necessitating the adoption of advanced technologies such as machine learning (ML) for enhanced detection capabilities. This research paper presents a comparative analysis of various machine learning algorithms—Naive Bayes, Logistic Regression, SGD Classifier, XGBoost, Decision Tree, Random Forest, and MLPClassifier—focusing on their effectiveness in phishing email classification. By evaluating key performance metrics such as accuracy, validation accuracy, classification reports, and confusion matrixes, the study identifies the strengths and weaknesses of each algorithm. The results highlight the SGD Classifier and Logistic Regression as the most robust models, exhibiting high generalization capabilities suitable for practical applications. Conversely, models like XGBoost and Decision Tree showed possible overfitting, indicating a need for further tuning. The findings provide valuable insights for cybersecurity professionals, aiding in the selection and implementation of effective phishing detection systems. Future research should explore hybrid models and advanced feature extraction techniques to further enhance detection accuracy and robustness
Equipping teachers for differentiation
No full textTeachers are expected to tailor their instruction to accommodate differences between students, in order to give all learners the opportunity to develop to their full potential. This is also called differentiated instruction (DI), and it requires teachers to continually monitor students’ progress towards the lesson goal and adjust their instruction accordingly. This is a complex skill for many teachers that they do not always implement. The goal of this dissertation was to research how educators can be supported in providing differentiation. In Study 1 we observed secondary school teachers with strong DI skills in their classrooms and interviewed them afterwards to analyse their decision-making processes. The findings of this study informed the design and implementation of a professional development program in Study 2, which was based on a whole-task approach where the participant teachers work on authentic tasks that require the integration of skills and knowledge, increasing the likelihood of transfer to practice. The impact of this professional development program on teachers’ DI practices was examined in Study 3, both from the teacher and student perspective. Results from both perspectives showed that all teachers made more intentional adaptations. To explore whether the findings of the first three studies were transferable to another context, we shifted the focus to students with hearing and/or communicative impairments (HCI). Instead of a single teacher, often a team of professionals has joint responsibility for HCI students’ education. In Study 4, we researched what providing DI and interprofessional collaboration in the HCI context looks like, by conducting a systematic literature review and an exploration of practice through focus group sessions. We found that there are four phases (preparing a lesson series, preparing a lesson, enacting a lesson, and evaluating the lesson) and five principles (work in a goal-oriented way, monitor continually, challenge all students, adapt the instruction, and stimulate self-regulation) that could be identified in DI across various educational contexts. The four phases and five principles can be used as a framework to support educators in improving their DI skills and thus making good, informed decisions that allow all their students to realise their learning potential
Artificial Intelligence in surgery:predicting outcomes and measuring performance
No full textArtificial Intelligence (AI) holds promise for surgery by supporting decision-making, personalizing treatment strategies, and enhancing performance evaluation. As data is collected before, during, and after surgical procedures, AI can be applied across the entire perioperative spectrum. This thesis investigated how AI can perform preoperative prediction of surgical outcomes and postoperative assessment of surgical performance. Six studies form the foundation of this work. The first part focused on predicting anastomotic leakage after rectal cancer surgery and the duration of gallbladder surgery. In the first study, a deep learning approach was developed to automate the measurement of pelvic dimensions in magnetic resonance imaging scans. The model performed at a human level, enabling an objective and standardized assessment of the dimensions relevant to surgical difficulty. In the second study, multimodal machine learning models were developed to predict anastomotic leakage after rectal cancer surgery preoperatively. Pelvic inlet width, interspinous distance, and distance to the anorectal junction were identified as independent predictors. While the predictive performance of the models remained moderate, the study highlighted the relevance of pelvimetry in clinical decision-making. In the third study, an adaptive framework was created to predict the procedure duration of laparoscopic cholecystectomy based on preoperative factors and intraoperative difficulty. The procedure duration was independently associated with the level of expertise, history of cholecystitis, hydropic gallbladder, and surgical difficulty. While the model's predictive value was modest, the study highlighted the importance of intraoperative difficulty for dynamic, personalized surgical planning.The second part addressed postoperative performance assessment through the analysis of electrosurgical device usage. An energy dashboard was introduced to visualize information extracted from energy generator data, providing a basis for feedback and benchmarking. The pilot study demonstrated significant differences between surgeons in total usage, turn-on count, and amount of applied energy. Building on this, a machine-learning-based algorithm was developed to automatically detect device-induced bleeding in laparoscopic diaphragmatic hernia repair. Although precise detection proved challenging, the algorithm provided a novel performance marker for suboptimal energy application. Finally, surgical phase recognition in diaphragmatic hernia repair videos was implemented to contextualize electrosurgical events within procedural steps. The deep learning methodology can accurately recognize the course surgical phases, but requires further development to distinguish the different dissection parts.Collectively, this thesis laid the groundwork for preoperative and postoperative tools to support surgery through standardized analysis of clinical, imaging, and intraoperative data. By quantifying anatomical constraints, risk factors, and device usage, these studies facilitate objective decision-making and performance evaluation. AI played a crucial role in converting multimodal data into interpretable metrics for risk assessment, surgical planning, and skill development. While current predictive models and assessments need further refinement for clinical use, they demonstrate a scalable framework that enhances surgical expertise with context-aware insights, ultimately improving patient treatment and continuous development
Gaussian or plane? Both:: Semantic-driven voxel representation for LiDAR–inertial odometry
No full textAccurate LiDAR-inertial odometry (LIO) highly depends on the geometric fidelity of the underlying environment representation. We explore the new and interesting research direction of integrating semantic segmentation models into metric odometry algorithms to enrich their representational capacity. Specifically, this letter proposes a semantic-driven hybrid voxel representation in which an off-the-shelf 3D segmentation network assigns every voxel to either a planar or nonplanar class, using planar and Gaussian representations, respectively. Consequently, a hybrid scan matching strategy is presented using class-specific residual models that are tailored to the distinct error statistics of each surface category. The scan matcher is embedded within an Iterated Extended Kalman Filter (IEKF) for odometry and mapping. We evaluate our method on diverse platforms and environments, and show improved localization accuracy across various indoor and outdoor scenarios, while maintaining real-time performance
Engineering Morphologies of Metal-Based Colloidal Assemblies via Colloid Jamming at Liquid–Liquid Interfaces
No full textAssemblies structured from colloids exhibit promise for advanced applications, including photonic devices, electrochemical energy storage systems, and catalytic supports. Despite observing diverse morphologies, a comprehensive understanding of the underlying formation mechanisms remains elusive. In this work, it is proposed that the coordination interactions between metal sulfide nanoparticles (MS NPs) and fluorosurfactants at the droplet interface influence the morphology of assemblies during the evaporation-induced self-assembly in droplet microfluidics. Systematic studies suggest that coordination strength significantly influences the morphology of assembly. The interfacial interactions can be effectively eliminated by coating the MS NPs with a SiO2 shell, forming metal sulfide@SiO2 nanoparticles (MS@SiO2 NPs). Furthermore, it is demonstrated that the morphology of assemblies can be engineered via tuning MS NPs concentration under coordination regulation. With interfacial jamming, core-shell or homogeneously distributed binary colloidal assemblies are constructed. These findings highlight the importance of coordination interactions and concentration in shaping colloidal assemblies during evaporation-driven self-assembly in surfactant-stabilized microdroplets. This insight provides a foundation for designing functional materials with controlled architectures for applications in catalysis, plasmonics, and porous materials.</p
Input-Output Extension of Underactuated Nonlinear Systems
No full textThis letter proposes a method to integrate auxiliary actuators that enhance the task-space capabilities of commercial underactuated systems leaving internal certified low-level controller untouched. The additional actuators are combined with a feedback-linearizing outer loop controller, enabling full-pose tracking. We provide the conditions under which legacy high-level commands and new actuator inputs can be cohesively coordinated to achieve decoupled control of all degrees of freedom. A comparative study with a standard quadrotor–originally not designed for physical interaction–demonstrates that the proposed modified platform remains stable under contact, while the baseline system diverges. Additionally, simulation results under parameter uncertainty illustrate the approach’s robustness.</p