58622 research outputs found
Sort by
Deep predictive coding with bi-directional propagation for classification and reconstruction
Get PDFPredictive Coding (PC) has emerged as a prominent theory underlying information processing in the brain. The general concept for learning in PC is that each layer learns to predict the activities of neurons in the previous layer, which enables local computation of error as well as in-parallel learning across layers. Deep Bi-directional Predictive Coding (DBPC) is proposed here as a new learning algorithm that enables neural networks to simultaneously perform classification and reconstruction tasks using the same learned weights. Building on existing PC approaches, DBPC supports both feedforward and feedback propagation of information. Each layer in the network trained using DBPC learns to predict the activities of neurons in the previous and next layers, enabling the network to simultaneously perform classification and reconstruction tasks using feedforward and feedback propagation, respectively. DBPC also relies on locally available information for learning, thus enabling in-parallel learning across all layers in the network. DBPC enables the training of both fully connected networks and convolutional neural networks. The classification accuracies of DBPC on the MNIST, Fashion-MNIST, and CIFAR-10 datasets (99.58%, 92.42%, and 74.29%, respectively) exceed those of well-established PC-based benchmark approaches (including FIPC3 and iPC) and are competitive with state-of-the-art Error-Backpropagation-based methods (including ResNet and DenseNet) on MNIST, Fashion-MNIST, and EuroSAT datasets. Importantly, DBPC achieves these results using significantly smaller networks for MNIST, Fashion-MNIST, and CIFAR-10 datasets (0.425, 1.004, and 1.109 million parameters), and every representation estimated in DBPC can be used for the reconstruction of inputs. The significant benefit of DBPC is its ability to achieve this performance using locally available information and in-parallel learning mechanisms, which results in an efficient training protocol. Overall, we demonstrate that DBPC is a much more efficient approach for training networks that can perform both classification and reconstruction simultaneously
Developing mathematical innovators to bridge research and impact in data-driven applications
Get PDFThe effective deployment of mathematical sciences in interdisciplinary, data-driven research remains a challenge for universities. Despite underpinning the UK’s digital research capability, mathematics often fails to translate into real-world impact, in part due to the lack of structured roles for researchers working at the interface of academia, industry, and government. This paper presents the Mathematical Innovation Research Associate (MIRA) model, developed at the Institute for Mathematical Innovation, University of Bath. The model formalises the role of PhD qualified mathematical scientists as embedded research professionals. MIRAs apply techniques from applied mathematics, statistics and machine learning to projects spanning health, engineering, social sciences and public policy. Their time is allocated across externally funded work, with flexible deployment and cost recovery mechanisms ensuring sustainability. We describe the model’s operational, financial and professional development structures. Since 2018, the MIRA team has supported over 60 projects and contributed to £27 million in research income. The model is now being adopted nationally via the MInDS programme, which supports institutions to establish similar roles. The MIRA model offers a scalable, transferable framework for embedding mathematical capability into research ecosystems, bridging the gap between academic expertise and applied, interdisciplinary impact
Hydrate-Based H2 Storage with Porous Materials as Heterogeneous Promoters:State of the Art and Challenges
Get PDFClathrate hydrates, which can store hydrogen inside crystalline, ice-like structures, have great potential for hydrogen storage. However, kinetic and thermodynamic promoters are often needed to improve the formation rates and stability ranges. Porous materials exhibit significant potential for hydrate-based hydrogen storage by modulating the kinetics, stability, and storage capacity, unlocking substantial application prospects. This review systematically elucidates the critical mechanisms through which porous materials influence hydrogen hydrate behavior, with a comprehensive analysis of the synergistic roles of material properties and engineering operation conditions. Material properties include the nano-confinement effect, which markedly enhances hydrate formation, optimized pore and particle sizes that increase contact area, functionalized surfaces and rough structures that improve nucleation and stability, and moderate hydrophobicity that enhances gas–water contact. Engineering operation conditions involve maintaining suitable temperatures and pressures to ensure stable hydrate formation, uniform spatial layouts to optimize gas diffusion, and water saturation control to boost reaction efficiency. The review further summarizes the application characteristics of various porous materials, including carbon-based materials (e.g. activated carbon), inorganic materials (e.g. silica), organic porous polymers (e.g. polyurethane foam), and hybrid materials (e.g. metal–organic frameworks), evaluating their respective strengths, limitations and suitability. Multiscale insights highlight the macroscopic focus on hydrate formation within high-pressure reactors, the mesoscopic emphasis on optimizing particle surface reactions, and the microscopic attention to confined hydrate growth within pore structures. Future research should prioritize the refinement of nanopore architectures, the development of advanced hydrophilic/hydrophobic materials, the enhancement of reactor designs, and the integration of thermal management and kinetic optimization to propel hydrogen hydrate storage technology toward practical implementation
The State of the Art in Foreign Policy Analysis: An Introduction
Get PDFThis introductory article outlines the purpose and scope of a special issue dedicated to assessing the current state of Foreign Policy Analysis (FPA). We highlight FPA’s evolution as a subfield within International Relations, particularly emphasizing its distinctive focus on agency and the interplay between domestic and international factors in shaping foreign policy. We also introduce the special issue’s contributions to the literature from emerging and established scholars in exploring theoretical, methodological, and empirical innovations in FPA. We conclude that by embracing methodological pluralism and diverse units of analysis, the special issue showcases how FPA continues to enrich our understanding of international politics while providing ideas for promising directions for future research
Construction of metal-organophosphate (MOPMs) interlayers for preparing high-performance polyamide composite nanofiltration membranes
Get PDFAchieving an ideal balance between water permeance and separation selectivity remains a significant bottleneck in advancing nanofiltration membrane for dye/salt separation. This work reports the use of metal–organic phosphate membranes (MOPMs) as an interlayer to regulate the retention and diffusion of piperazine (PIP) monomers on the substrate, thereby optimizing the thickness and bacteriostasis properties of the polyamide layer. Under optimal modification conditions with 0.2 mg/mL PA, the water permeance of the MOPMs-0.2-IP considerably elevated to 42.18 L m−2h−1bar−1, approximately double that of the controlled membrane, while maintaining comparable dye/salt separation performance (Congo red >99.9 %, NaCl: 18.7 %, MgCl2: 12.4 %). Noticeably, compared with existing nanofiltration membranes, MOPMs-0.2-IP exhibits excellent permeability and separation capabilities. This work provides valuable insights into optimizing polyamide layers through the use of intermediate layers, advancing the development of high-performance nanofiltration membranes.</p
“It’s Like All of My Senses and My Body Become More Awakened”:Autistic Adult’s Experiences of Attending Live Music
Get PDFBackground: Up to 94% of Autistic people have sensory responsivity differences, associated with experiences that can range from being distressing to highly pleasurable. Despite the importance of live music in fostering social inclusion, many venues and events are not inclusive, creating barriers for Autistic people. Given the challenges Autistic people face in accessing live music events, our study aimed to explore their experiences to improve inclusivity, using a critical realism epistemological approach. Methods: We conducted a qualitative study with 16 Autistic adults aged 21-52 through online focus groups (n = 13), individual interviews (n = 1), and email exchanges (n = 2), allowing for spoken or typed communication. Each focus group lasted 1 hour, and we analyzed the data using reflexive thematic analysis, following good practice guidelines. Results: We developed four main themes and two subthemes: (1) “This is a military operation”—Planning to manage uncertainty and overwhelm; (2) “Hating a crowd and loving a crowd at the same time”—Social connection; (3) The duality of an intense sensory environment; (3a) “The music, the vibrations bring out the colors … in my mind”—Immersive sensory joy; (3b) “My brain is screaming at me”—Sensory overload; and (4) “I think the biggest difference, always, is the staff”—Combating stigma and creating safe spaces. Conclusions: Our study highlights the positive and negative aspects of attending live music for Autistic people, offering actionable recommendations for inclusivity. Key suggestions include providing advance information, earplugs/quiet spaces, minimizing crowd exposure, and ensuring staff are trained to support Autistic attendees. These measures can help create accessible, enjoyable live music experiences, fostering social connections and reducing isolation for Autistic people.</p
Climate Anxiety in Perspective: A Look at Dominant Stressors in Youth Mental Health and Sleep
Get PDFThere is growing evidence that climate anxiety is associated with significant effects on the mental health and wellbeing of young people. However, the relative importance of climate anxiety for young people's mental health has hitherto been unclear, as climate anxiety has largely been studied in isolation from other common stressors. This study sought to contextualize the significance of climate anxiety for the mental health of UK young adults relative to other concurrent psychological stressors. We surveyed university students (N = 461) and a general population sample aged 18–25 (N = 400). The results showed that while climate anxiety was significantly associated with poorer mental health and worse insomnia when examined alone, this association became nonsignificant or greatly diminished when other stressors were considered. Loneliness was found to be the most important predictor of mental health, and financial anxiety the most important predictor of insomnia severity. The findings suggest that climate anxiety, while concerning, may not be an especially dominant factor in young people's mental health. Our research highlights the need to consider the broader context of young people's lives, and the complex interplay of various psychological stressors, in efforts to map pathways between climate change and mental health
Three-day blues after ecstasy/MDMA use:Evidence from a longitudinal and daily analysis in the European nightlife scene
Get PDFBackground: There is a lack of understanding of the nature of the post-acute affective response in the days after ecstasy/3,4-Methylenedioxymethamphetamine (MDMA) use and whether this is associated with ecstasy/MDMA use or circumstantial factors. In the three days following ecstasy/MDMA use, we evaluated whether a drop in mental well-being is observed and can be related to ecstasy use. Methods: Data for this study were obtained from a longitudinal and momentary analysis in the European nightlife scene (ALAMA study). Using ecological daily assessment, participants were asked to complete a daily 3-minute questionnaire for 35 days. Young adults (age 18–34) from the United Kingdom (n = 120) and the Netherlands (n = 124) who use ecstasy/MDMA were recruited in the nightlife scene and using social media campaigns. Substance use, psychological well-being and pathology, sleep quality, harm reduction behaviours, and socio-demographics data were collected digitally through a smartphone app. Results: Participants reported on average a significant drop in mental well-being in the three days following ecstasy/MDMA use (B=-0.14, SE=0.04, p < .001) even when accounting for other substance use, socio-demographics, applied harm reduction strategies, measures of depression, anxiety and sleep quality. For commonly used substances other than ecstasy/MDMA and cocaine, no significant associations with mental well-being in the three days following their use were found. Conclusions: A drop in mental well-being in the three days following ecstasy/MDMA use was associated with ecstasy/MDMA use, in addition to other factors such as (co-)use of other substances, especially cocaine, sleep duration and quality in the days following use, and baseline levels of depression and anxiety.</p
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇