19684 research outputs found
Sort by
Cognitive feature evaluation for disease progression in dementia and its precursors using feature selection
No full textPurpose: Dementia is a condition with symptoms of memory decline, cognitive impairment, and difficulties in language and problem-solving, among others. Early screening of dementia conditions such as Alzheimer’s disease (AD) is fundamental for quick intervention and disease management. Currently used neuropsychological assessments are time-consuming as they contain many elements and require critical resources which are not always available. Other pathological assessments are invasive and not cost effective, hence identifying cognitive features for different dementia sub-groups during progression of the condition is crucial for clinicians. This study investigates this problem using a cost-effective data driven approach. Methods: Using real cases and controls from the Alzheimer’s Disease Neuroimaging Initiative data repository (ADNI) who have undergone the Alzheimer’s Disease Assessment Scale-Cognitive 13 (ADAS-Cog), we conduct a feature-feature assessment together with Permutation Feature Importance (PFI) and machine learning algorithms to derive influential cognitive features for specific dementia groups from baseline diagnosis up to 36 months. Results: Feature-feature analysis showed correlations between memory tasks such as Word Recall, Delayed Word Recall, and Word Recognition across both CN-MCI and MCI-AD groups. In contrast, low correlations for Naming, Command, and Ideational Praxis suggest they tap into distinct DSM-5 domains thus making them ideal for early screening. In addition, PFI results showed that Delayed Word Recall emerged as a top cognitive marker of progression in early stages, while Orientation gained prominence later thereby reflecting a shift toward executive and attentional decline. Conclusions: The results of this study identified important relationships between cognitive features in the ADAS-Cog and provide a clear example of the value of data-driven machine learning approaches in the identification of markers that indicate disease progression in dementia.</p
Literary Heritage:Lessons from the Coronavirus Pandemic
No full textLiterary Heritage examines the literary heritage sector in the post-pandemic moment. The book argues that this is a unique time for literary heritage management and demonstrates that the key to understanding it is an analysis of the transformations that took place because of the Covid-19 pandemic. Through an analysis of literary heritage sites across the UK's four nations, this study provides an overview of practice from sites managed by national organisations as well as independent museums. Presenting a quantitative and qualitative overview of the challenges faced by the sector in the wake of the pandemic, Rudrum and Williams explore the innovations literary heritage organisations initiated in response. The book displays the wealth of ingenuity that was on display during this trying moment for the sector. It also looks forward to the new normal in the industry: a move towards the outdoors, increased use of online engagement, and creative arts and community programming that brings the literary past to the political present. Featuring interviews with 16 heritage practitioners, the book shares examples of best practice in the hope that lessons will be learned from the enforced closures prompted by the pandemic. Literary Heritage will be of great interest to academics and students working in Heritage Studies, Museum Studies, and English Literature. It will also appeal to a broad readership of cultural heritage professionals.The Open Access version of this book, available at http://www.taylorfrancis.com, has been made available under a Creative Commons Attribution-Non Commercial-No Derivatives (CC BY-NC-ND) 4.0 license.</p
ΑI-Based Foreign Language Learning Tools Effectiveness
No full textIn this survey several studies are presented on foreign language learning assisted by AI-based tools/chatbots in a classroom setting, mostly using 5G-enabled smartphones. The focus of these studies is the qualitative and quantitative assessment of the effectiveness of the tools. Pre-tests and post-tests are frequently used to quantitatively assess student improvement with and without the use of AI-tools in several areas of language learning (vocabulary, grammar, syntax, listening, speaking, writing, etc.). A control student group is used in many cases as a baseline for assessing improvement. Furthermore, a qualitative questionnaire is filled in by the students in most of the studies to assess the strengths and weaknesses of the tools from the perspective of the users. It is found that the use of AI-based tools leads to significant improvement in learning progress.</p
Evaluating Ballistic Surfaces in Converted Firearms Using X-ray Computed Tomography to Establish Mode of Use and Identification
No full textThe Application of Criminal Behaviour and Forensic Evidence for the Rapist Profiles in Thailand
Get PDFBiological Systems as Foundries for the Novel Manufacture of Nanotechnology via DNA Origami
No full textA hybrid theoretical–numerical–experimental framework for robust health monitoring of thin-walled hollow composite members using guided waves
No full textThin-walled hollow composite members (HCM) are extensively employed in aerospace and automotive industries due to their high strength-to-weight ratio and design flexibility. This study introduces a hybrid -numerical–experimental framework for robust detection and characterisation of barely visible damage in HCM using guided waves (GW). It focuses on assessing surface abrasion and hairline cracks, two common yet challenging damage types encountered in the field. A semi-analytical finite element (SAFE) formulation is developed for the dispersion analysis alongside numerical simulations using finite element software COMSOL Multiphysics®, and experimental validation is performed to ensure accurate and reliable results. The study focuses on GW propagation and scattering behaviour under varying damage scenarios, exploring the effects of damage size, position, and its offset on wave features. Parametric analyses show significant variations in wave characteristics such as group velocity, amplitude, and mode features. A waveform and statistical approach incorporating continuous wavelet transform (CWT) and energy enables precise damage classification. Results show that abrasioninduced damages cause substantial changes in GW features in terms of DIs and statistical parameters, while hairline cracks marginally affect the damage indices and wave features, aiding in distinguishing between different damage types. These findings contribute to the development of robust damage identification algorithms for structural health monitoring, providing valuable insights for optimising the maintenance and performance of composite structures in critical engineering environments,ensuring safety and operational efficiency
Automating power plant and power electronic controller tuning for enhanced grid stability
No full textThis paper presents the development of an automated model for tuning power plants and power electronic controllers in electrical power systems, addressing challenges posed by the integration of renewable energy sources and the resulting reduction in grid inertia. The traditional manual tuning process, complicated by proprietary black-box models from Original Equipment Manufacturers (OEMs), is time-consuming and requires high expertise and, as of September 2022, it is no longer allowed by the National Grid. This work proposes a generic open-source model that emulates OEM systems, facilitating grid code compliance through automated tuning. The model integrates PowerFactory simulations with Python scripting and a Windows Forms interface, optimising control parameters such as proportional gain (Kp) and integral gain (Ki) using machine learning algorithms. The methodology includes a detailed literature review, robust research design, and validation of a model power system. Results demonstrate significant improvements in tuning efficiency and system response, offering a scalable solution for various power plants, enhancing the integration of renewable energy, and promoting grid stability.</p
A Survey on Model Repair in AI Planning
No full textAccurate planning models are a prerequisite for the appropriate functioning of AI planning applications. Creating these models is, however, a tedious and error-prone task - even for planning experts. This makes the provision of automated modeling support essential. In this work, we differentiate between approaches that learn models from scratch (called domain model acquisition) and those that repair flawed or incomplete ones. We survey approaches for the latter, including those that can be used for domain repair but have been developed for other applications, discuss possible optimization metrics (i.e., which repaired model to aim at), and conclude with lines of research we believe deserve more attention.</p
Access Repository Dashboard
Do you manage Huddersfield Research Portal? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!