18624 research outputs found
Sort by
Discriminating malware families using partitional clustering
No full textMalware, malicious software designed to compromise device security, is crafted by expert software engineers and distributed through a specialized black markets. Identifying malware families within daily feeds remains a significant challenge for internet security firms. Industry-standard Yara rules, based on regular expressions, are prone to failure due to malware evolution. This thesis presents an alternative approach leveraging malware clustering. By clustering malware samples based on dynamic analysis features, Yara scans can efficiently pinpoint known families, but unrecognized samples signify potential new variants, earmarked for further scrutiny by analysis teams. This process diminishes the necessity for individual sample scans, thereby streamlining operations and lightening the analysis team’s workload. This research evaluates the partitional clustering algorithm for improved handling of sparse malware features, setting it against the following traditional algorithms K-Means, Agglomerative Clustering, DBSCAN, and Spectral Kmeans Clustering. Each algorithm is evaluated, with a focus on their efficacy clustering performance: KMeans optimizes for homogeneous variance across n groups; Agglomerative Clustering scales for large datasets via connectivity matrices; DBSCAN discriminates clusters based on density metrics; and Spectral K-means Clustering employs affinity matrix-based low-dimensional embedding prior to clustering. The contribution of this thesis include a comprehensive performance comparison of the partitional clustering algorithm against Hierarchical, Densitybased, Spectral K-means, and K-Means algorithms; enhancement of the partitional clustering algorithm for sparse data; an in-depth evaluation of features extracted from Application Programming Interface call parameters and Domain Name System queries executed by malware; and the development of countermeasures against malware’s anti-analysis tactics. The research utilizes a real-world malware dataset sourced from abuse.ch 1 [1]. Empirical results demonstrate the superior performance of the partitional clustering algorithm over traditional clustering techniques in the majority of tests conductedMasters of Researc
Experiments with a dark pedagogy : learning from/ through temporality, climate change and species extinction (::: and Ghosts)
Get PDFIn this article, we experiment with a form of dark pedagogy, a pedagogy that confronts haunting past
Unveiling soil-vegetation interactions : reflection relationships and an attention-based deep learning approach for carbon estimation
No full textEstimating soil organic carbon (SOC) from satellite imagery, particularly in areas with both bare soil and vegetation, poses significant challenges. Traditional approaches often overlook the complex interactions between soil and vegetation. Addressing this gap, our study introduces an innovative method that leverages novel correction of hyperspectral reflections to adjust for vegetation levels, enhancing SOC estimation accuracy. Moreover, we propose an attention-based deep neural network that dynamically prioritizes spectral features crucial for SOC prediction. This mechanism significantly improves the model's ability to detect significant features for accurate SOC estimation. Comparative experiments with traditional models on a benchmark dataset demonstrate our method's effectiveness in reducing vegetation influence and accurately estimating SOC across mixed landscapes. Our findings represent a notable advancement in SOC estimation from satellite imagery, highlighting the potential of advanced learning-based techniques with attention-driven feature weighting for SOC estimation. © 2024 IEEE
A simplified optimal switching sequence model predictive control without weighting coefficients for t-type single-phase three-level inverters
No full textThis article proposes a simplified optimal switching sequence model predictive control (OSS-MPC) without weighting coefficients for off-grid T-type single-phase three-level inverters. To eliminate the tedious weight-tuning process when constructing the switching sequences, the different effects of redundant small voltage vectors on the upper and lower dc bus capacitors' voltages are considered. Each switching sequence consists of two symmetrical redundant small voltage vectors with opposite effects on dc bus capacitors' voltages so that the neutral point voltage balance of the inverter can be achieved by tuning the dwelling time of two redundant small voltage vectors with opposite effects. According to the different effects of the switching sequence, the candidate switching sequences can be obtained by a simple division of the switching sequences through the reference. Then, the OSS can be selected by minimizing a cost function, which considerably simplifies the optimization process. Finally, the effectiveness of the proposed simplified OSS-MPC algorithm is experimentally evaluated based on a laboratory prototype in terms of code execution time, steady-state performance, transient performance, and neutral point potential balance performance. © 2015 IEEE
What we’re taking through the portal: how our experiences of remote teaching as parent-educators during COVID-19 impacted our practice
No full textThis chapter investigates the perspectives and learning that we as parent educators developed during the COVID-19 pandemic and how this impacted our pedagogies. In 2020 and 2021 our teaching shifted from face-to-face to online delivery while we facilitated our children’s remote schooling. The disruption caused by the pandemic was a portal-like opportunity to rethink our approaches to learning and teaching. Our collaborative reflections on our changing ideas about learning and pedagogies form the basis of this chapter. We adapted conceptual frameworks including narrative inquiry and the Indigenous notion of yarning to develop a collaborative narrative inquiry approach. To examine the complexities of interconnected individual, social and cultural elements, we draw on a theoretical framework that combines Bronfenbrenner’s Ecological Model and Antonovsky’s Salutogenic approach. The chief findings of the research showed that a focus on inclusive pedagogies including student voice and agency, was not just desirable for learners but was imperative to ensure learner engagement. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2024
Fostering social justice and inclusion in teacher education in global contexts
No full textThis chapter introduces key concepts such as social justice, inclusion, equity, diversity, and effective learning for all and the implications for teacher education. It provides an outline of how the contributing authors explore, interrogate, and critique educational practices and structures that contribute to educational disadvantage or exclusion, aiming to facilitate inclusive mindsets in the field of teacher education. Critical perspectives of diversity and equity in teacher education are investigated through an array of traditions and methodologies that interrogate educational issues from a political, cultural, structural, and social perspective. Barriers and enablers facing inclusion in teacher education are explored and pedagogical approaches to address equity, diversity and inclusion in teacher education are cultivated and critiqued. Attention is directed towards building the capacities of teacher education stakeholders to understand and respond to teaching and learning contexts through a standpoint of equity and justice. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2024
TRL-SN : trajectory representation learning with spatial networks for travel mode identification
No full textTravel mode identification is an important research area in trajectory mining. It aims to identify different travel modes based on the analysis of travel trajectory data. However, the existing research has mostly focused on the intrinsic attributes of trajectories, thereby overlooking the road network information associated with them and lacking consideration of the spatial interactions among trajectories. We proposed a trajectory representation learning with spatial networks framework (TRLSN) for travel mode identification. Firstly, we employed map matching and graph learning techniques to project the trajectory onto the road network, achieving an accurate graph model union of the trajectory and the road network. Then, we designed a road network information interaction model based on graph attention networks that generate road segment vector representations by capturing the spatial interactions of trajectories. Subsequently, we designed a trajectory embedding encoding model based on Transformer to mine the periodic patterns of trajectories, obtaining more effective trajectory vector representations. Finally, we identified travel modes through a discriminator. We conducted comparative experiments on a real dataset to verify the effectiveness of this method. © 2024 IEEE
Motif-induced subgraph generative learning for explainable neurological disorder detection
No full textThe wide variation in symptoms of neurological disorders among patients necessitates uncovering individual pathologies for accurate clinical diagnosis and treatment. Current methods attempt to generalize specific biomarkers to explain individual pathology, but they often lack analysis of the underlying pathogenic mechanisms, leading to biased biomarkers and unreliable diagnoses. To address this issue, we propose a motif-induced subgraph generative learning model (MSGL), which provides multi-tiered biomarkers and facilitates explainable diagnoses of neurological disorders. MSGL uncovers underlying pathogenic mechanisms by exploring representative connectivity patterns within brain networks, offering motif-level biomarkers to tackle the challenge of clinical heterogeneity. Furthermore, it utilizes motif-induced information to generate enhanced brain network subgraphs as personalized biomarkers for identifying individual pathology. Experimental results demonstrate that MSGL outperforms baseline models. The identified biomarkers align with recent neuroscientific findings, enhancing their clinical applicability. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025
ESG assessment methodology for emerging technologies : plasma versus conventional technology for ammonia production
Get PDFEnvironmental, social and governance (ESG) criteria demand that enterprises should not be assessed solely on their financial performance, but also on their environmental, social, and governance performance. This numerical assessment of ESG criteria enables them to be evaluated with the consideration of other financial issues of enterprises' performance and thereby guides financial investments into environmentally and socially responsible firms. ESG, however, solidifies the continuance of conventional technologies but can potentially disadvantage emerging technologies. This study is the first to forecast the ESG potential of emerging chemical technologies. The Morgan Stanley Capital International (MSCI) rating system is applied to one of the top 3 global chemical processes. Ammonia (NH3) is produced via the Haber-Bosch (HB) process, which needs a huge fossil fuel input and high energy consumption, leading to a significant contribution to carbon dioxide (CO2) emissions. In contrast, the ESG assessment rates emerging plasma technology and its spearhead companies that lead innovation and development in this field, which provide the benefits of being a clean, sustainable alternative for green NH3 production. Five different plasma-technology companies are considered, with the technology readiness level (TRL) ranging from 3 to 9. These are compared to five different conventional HB companies. We examine the final ESG result of the plasma technology companies, exploring their environmental advances and social viability. In this study, five different themes were selected, including eleven issues, to measure the plasma-technology company's management related to ESG risks and opportunities. © 2025 RSC
Race(ing) social work in Australia : three critical recognitions for dismantling racism
No full textThis chapter takes as its starting point the paradoxical nature of Australian social work practice wherein the discourse of anti-racist (and anti-oppressive) practice co-exists with the reality of pervasive racism in a whitewashed profession. As a profession whose core is working with and advocating for historically oppressed and marginalized groups, it follows that proclamations of anti-racist practice predominate. The main contention in this chapter is that while the discourse of anti-racism in Australian social work practice continues, in reality, the crucial foundational work needed to dismantle racism within the profession has still not been done. Using a framework of "racism as violence," it argues that central to this foundational work are three critical recognitions of racism as: (1) systemic violence that broadly underpins social work practice; (2) ideological violence of an invisible White, western worldview and White racial framework that continues to inform social work practice; and (3) structural violence embedded in "normalized" racist structures, policies, and practices that still exist at the core of social work practice and benefit racially privileged (White) people. In the end, this chapter is a call for explicit, intentional, and widespread analytical discussion of racism within the profession. These analyses must center and be informed by the "voice of color" as articulated in Critical Race Theory - i.e., by racially marginalized social workers with lived experiences of racism. Such analyses immanently advocate for a critically reflective, structurally focused anti-racism as the only way to dismantle racism in Australian social work. © Springer Nature Singapore Pte Ltd. 2024. All rights reserved