25096 research outputs found
Sort by
Dynamic multiobjective evolutionary algorithm based on a knee point driven Gaussian model
No full textThe file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.Knee point driven dynamic multiobjective evolutionary algorithms have been proven to be promising, however, the predicted knee points with worse quality may mislead the evolution. One of reasons for obtaining worse knee points is that previous methods focus on various prediction models, but neglects the effective strategy for exploring knee points according to the characteristics of decision vectors, falling into incorrect prediction direction. To address this issue, dynamic multiobjective evolutionary algorithm based on a knee point driven Gaussian model is proposed, termed KG-DMOEA. Once an environmental change is detected, knee points of last environment are extracted, and decision variables are classified into convergence and diversity related ones. Following that, knee point exploration strategy is developed to estimate new knee points in terms of historical knee points and convergence related variables. These estimated ones are further introduced to build the Gaussian function as generative model, and diversity related variables are utilized to update the generative model, with the purpose of overcoming the challenge that these decision variables may change over time. Based on this model, an initial population is produced at new time, and the model is also updated during evolutionary process to generate offspring individuals, speeding up the convergence. The intensive experiments demonstrate that the proposed KG-DMOEA has promising computational efficiency and performance in solving dynamic multiobjective optimization problems, outperforming several state-of-the-art DMOEAs
‘It’s about having a regular conversation about their academic process’: evaluating embedded personal tutoring for first year students
Get PDFopen access articleThis article offers a case study of curriculum-embedded personal tutoring, taken from a Humanities department in a Post-92 university over a period of several years (2018-2022). A core first year taught unit was redesigned with personal tutoring at its core as part of a review of multiple areas of overlapping focus: supporting the transition to University (induction), helping students develop the skills required for degree-level study, building learning communities, encouraging cohort identity creation and improving attendance and engagement.
Using student and staff qualitative focus groups and analysing data around continuation gaps, the article demonstrates the benefits of a curriculum-integrated personal tutoring model
An IoT Architecture for Enhancing Safety in Supply Chain Systems
No full textSupply chain processes are used for controlling the delivery of goods from suppliers to consumers. With the globalization of the economy, delivery distances are on the increase and might involve thousands of kilometers across countries and even continents and different means of transport including land, water, and air transport. The task of the supply chain process is to guarantee the quality and safety of the goods at transportation time to prevent potential damage, deterioration, and decay. This requires real-time monitoring and observance of several conditions (e.g., temperature and humidity) to which the goods are exposed. To address the problem, this paper discusses the implementation of an IoT system and relevant metrics for monitoring, in real-time, the conditions that goods face during transportation. In the prototype that we have implemented, we use conventional low-cost cost widely available technologies, namely, temperature and light sensors connected to a Raspberry Pi device that sends the sensors' data to a database server. We then analyze the data with the help of a web application that we have also implemented
Modelling the Influence of Urban Morphology on Bikeshare Station Use: A Clustering Approach
Get PDFopen access articleDocked bikeshare schemes have proliferated across UK cities since the first scheme was introduced in 2010. These schemes have been widely adopted for their contributions to decarbonising transport, improving health, and enhancing connectivity through first and last-mile trips. As bikeshare expands to new cities, planners and operators increasingly require a localised understanding of the factors influencing bikeshare use. Urban morphology in UK cities varies widely, however, encompassing differences in street layouts, building design, accessibility, and land use. Meanwhile, industry bikeshare planning guidelines are often broad, without distinguishing between city size and character. These variations pose challenges for bikeshare scheme planning in different settings, emphasising the need for robust, data-driven models that are sensitive to urban context. This paper employs cluster analysis to classify urban areas within several UK cities, with the aim to understand the combined contextual urban factors that influence bikeshare use. This approach, rarely applied in micromobility research, offers a nuanced and unique methodological contribution. The cluster analysis distinguishes between types of residential neighbourhoods, which is a component less commonly incorporated within existing studies. With the data obtained, statistical analysis offers granular insights into the relationship between the built environment and docking station use. It is highlighted that denser residential neighbourhoods with favourable accessibility have consistent associations with trip generation, while accessible suburban neighbourhoods are more varied. The findings have implications for both initial planning and scheme expansion, relevant to station location optimisation, forecasting future demand, fleet size adjustment and integration with existing public transport networks
Getting Real: How Intersectionality Challenges the ‘One Size Fits All’ Approach to Global Mental Health
No full textopen access articleThis paper offers an intersectional perspective to enable inclusivity when addressing global mental health challenges, whilst also tackling social exclusion and vulnerability. Examples are given from purposefully selected countries to demonstrate the effectiveness of exploring issues through an intersectional approach. The authors conclude that a ‘one size fits all’ approach, when working towards universal mental health, is not the best approach and cannot provide a solution for all. They suggest the WHO, UN, governments and mental health agencies need to view global mental health differently to achieve alternative and better targeted solutions
A BiLSTM-IPPO approach for dynamic flexible flow shop scheduling with limited buffer capacity
No full textThis paper addresses the Dynamic Flexible Flow Shop Scheduling Problem with Limited Buffer Capacity (DFFSPLBC) by proposing Improved Proximal Policy Optimization with Bidirectional Long Short-Term Memory (BiLSTM-IPPO). We design state representations, action spaces and reward functions, and adapt the PPO framework to dynamically handle variable action dimensions. A Gantt-chart-based feature extraction method automatically captures comprehensive scheduling information, enhancing generalization. The bidirectional LSTM (BiLSTM) exploits forward and backward temporal dependencies to fully capture job-machine interactions. Simulation on benchmark instances of various scales demonstrates that BiLSTM-IPPO outperforms heuristic, reinforcement learning (RL) and deep reinforcement learning (DRL) approaches in makespan reduction, validating its effectiveness and scalability
Peroxidase-catalyzed Coloration for Fabric Design with Color Patterns
Get PDFThe file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.Biotechnology using enzymes has been explored in textile wet processing for the potential of reducing energy and water consumption, due to the use of the highly specific biocatalysts that can operate under mild temperature and neutral pH conditions. The current research study contributes to an understanding of the use of the enzyme peroxidase for textile coloration of wool fabrics as an alternative coloration method to using conventional dyestuffs. Peroxidases, belonging to the enzyme group of oxidoreductases, can catalyze oxidation of a wide range of colorless simple aromatic compounds as precursors to form polymeric colorants. This enzymatic coloration can be successfully applied to in-situ dyeing of wool fabrics at a low temperature through peroxidase catalysis of various precursors over a broad range of pH values to achieve a diverse color palette. To explore the potential of enzymatic coloration for fabric design, a woven wool base fabric was embroidered using computer-controlled embroidery machines with embroidery yarns of different fiber types and subsequently enzymatically dyed to create color patterns. Peroxidase-catalyzed coloration has the potential not only as an alternative coloration process to create design patterns of fabrics, but also for saving energy and preventing fiber damage during the dyeing process
Betwixt and Between II
No full textAn intermedia performance artwork for live solo movement and fixed media: single screen moving image, eight-channel fixed medium electraocoustic music, theatrical lighting.
First performance (peer-reviewed) 07.11.2025 at Sound/Image 2025 festival, University of Greenwich, Bathway Theatre, London.This exposition presents the next iteration of the authors’ long-standing collaborative explorations into the delicate and complex relationships between live-digital dance performance and acousmatic sound. Building on previous work, Betwixt & Between II seeks to ‘reflect forwards’ on the interplay between the body, media and sound. Many of the authors’ underlying concerns still resonate today, including ideas of intimacy, fragility, connection, and imagination. In our increasingly technologized world; opportunities to reconnect with bodies, images, and sounds reflectively and deliberately feel ever more significant. By challenging some of the normative traditions of combining movement, media, and music, particularly in terms of working generatively and spontaneously, this interactive and intuitive work presents bodies, media and sounds simultaneously as mutual sensuous entities
Disentangling the Effects of Firm‐Level Climate Risk and Capital Market Signalling: Evidence From Stock Price Informativeness
Get PDFopen access articleThis study examines the impact of firm-level climate risk on stock price informativeness (SPI) through the integrated lens of stakeholder–shareholder theory. Using a global unbalanced panel of 73,770 firm-year observations across 38 countries (2000–2020), we find that higher carbon emissions significantly reduce SPI, reflecting increased information asymmetry. Governance mechanisms, specifically board size, independence, tenure and nationality mix, consistently moderate this effect by enhancing disclosure and mitigating opacity. The negative relationship between emissions and SPI is strongest in common law countries and those with high institutional quality, where stricter enforcement and disclosure regimes heighten investor sensitivity to environmental risks. Additionally, we document that transparency in emission disclosure, financial risks and environmental liabilities is identified as a key channel through which firm-level climate risk affects market informativeness. Furthermore, higher SPI is associated with lower cost of capital, more efficient capital allocation and reduced crash risk. This study contributes novel insights to the climate finance literature by integrating firm-level governance factors with cross-jurisdictional analysis. Robustness checks, including placebo tests, alternative SPI measures and system GMM estimation, confirm the validity of our results and underscore the importance of institutional context in pricing environmental risk
Facilitators and barriers to early diagnosis of malignant plural mesothelioma (FILMM): Patients journey towards mesothelioma diagnosis, in England.
Get PDFPoster presentationBackground:
Prognosis with malignant plural mesothelioma (MPM), which is a rare form of cancer, is poor, yet evidence indicates a better chance of survival if earlier diagnosis is provided. Partly due to late presentation and diagnosis of malignant plural mesothelioma (MPM), the
survival rate in the United Kingdom is below the European average. Furthermore, there has been little attention to MPM patients’ experiences prior to diagnosis as available studies have focused on their lived experiences after diagnosis. This study therefore aims to identify the barriers and facilitators to MPM diagnosis and looks to understand the reasons for any variability in patients’ experiences of the pathway to diagnosis and proposed treatment plans.
Methods:
The theoretical basis for this study is the Model of pathway to treatment (MPT), which highlights four intervals (Appraisal; Help-seeking; Diagnostic; and Pre-treatment) along the pathway to diagnosis where a patient can experience ‘delay’ in obtaining a cancer diagnosis. This model was used to develop the interview topic guide.
Patients with confirmed MPM diagnosis were invited to take part in an in-depth, semi structured interview about their pathway to diagnosis. To ensure the recruitment of the targeted number of participants, participants were purposively recruited from two specialist
MPM outpatient clinics in England, of which, one of these clinics manages the second largest number of MPM patients in the country. A total of seventeen participants took part in the study.
The interview data were analysed using framework analysis. The MPT was used for the initial coding framework of the interview transcriptions. Common themes were then identified within each MPT interval.
Results:
Our findings identified barriers and facilitators within every interval along the MPM patients’ journey to diagnosis. Within the Appraisal and Diagnostic intervals, the presentation of vague symptoms that were mistaken for a less serious illness were found to
be a barrier. Health literacy regarding MPM appears to have an impact on how soon a patient sought help regarding their symptoms. Good health literacy by healthcare professionals (HCPs) appeared to facilitate placement on an MPM diagnostic pathway which
reassured patients about better potential outcomes.
Conclusion:
Earlier symptom recognition by both patient and HCPs including General Practitioners (GPs) can be used to target significant and avoidable delays along patients’ MPM diagnosis pathway, thereby promoting earlier diagnosis and treatment options
Data sources:
Access Repository Dashboard
Do you manage De Montfort University Open Research Archive? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!