30793 research outputs found
Sort by
Spatial–Temporal Deep Learning for Electric-Vehicle Charging Demand: An Exploratory Study of Graph Convolutional and LSTM Networks Performance
Get PDFElectric-vehicle (EV) charging is a localized, time-varying load that challenges distribution networks. This study offers practical insights into when spatial graph structure adds value beyond temporal context, utilizing real-world data and a transparent evaluation. We compare Long Short-Term Memory (LSTM) and Graph Convolutional Network (GCN) models for hourly EV-charging energy forecasting, based on 145,778 sessions recorded in Boulder, Colorado (2018–2023). After preprocessing and temporal alignment, temporal covariates (hour, day, month, year) and, when applicable, ZIP-code indicators were engineered. LSTMs were trained with 1 h and 24 h input windows, with or without ZIP features, and evaluated through 5-fold cross-validation. GCNs operated on hourly node–time tensors with a dense adjacency (no self-loops) and were trained on an 80/20 temporal split, using a 1% subsample for tractability. All models used Adam (lr = 0.005), early stopping, and ReduceLROnPlateau. Temporal context was the main driver of LSTM accuracy: 24 h inputs outperformed 1 h, while ZIP features improved only the shorter window. For GCNs, depth and node features shaped performance: a 2-layer GCN with ZIP features achieved the lowest RMSE (4.75 kWh), whereas a 6-layer GCN without node features reached the lowest MAE (2.46 kWh). Despite higher computational cost, GCNs captured spatial coupling effectively. Overall, 24 h LSTMs provide a strong and efficient baseline, while GCNs add value when spatial correlations are relevant. A hybrid GCN–LSTM architecture is a promising next step to jointly leverage spatial and temporal dependencies.Office of Research, Zayed University, through the Research Incentive Fund (Grant Number: R23079)
Search for -hadron decays to long-lived particles in the CMS endcap muon detectors
Get PDFA version of the article is available at arXiv:2508.06363v2 [hep-ex] (https://arxiv.org/abs/2508.06363). Comments: Replaced with the published version. Added the journal reference and the DOI. All the figures and tables can be found at https://cms-results.web.cern.ch/cms-results/public-results/publications/EXO-24-004 (CMS Public Pages). Report number: CMS-EXO-24-004, CERN-EP-2025-166. Journal reference: Phys. Rev. D 113 (2026) 012009. Submission history: From: The CMS Collaboration: [v1] Fri, 8 Aug 2025 14:48:27 UTC (995 KB); [v2] Sat, 17 Jan 2026 20:34:15 UTC (1,032 KB).Data Availability:
Release and preservation of data used by the CMS Collaboration as the basis for publications is guided by the CMS data preservation, re-use and open access policy [77]. DataAvailability, CMS data availability statement, 10.7483/OPENDATA.CMS.1BNU.8V1W.A search for long-lived particles originating from the decay of hadrons produced in proton-proton collisions with a center-of-mass energy of 13 TeV at the LHC is presented. The analysis is performed on a dataset recorded in 2018, corresponding to an integrated luminosity of 41.6 fb⁻¹. Interactions of the long-lived particles in the CMS endcap muon system would create hadronic or electromagnetic showers, producing clusters of detector hits. Selected events contain at least one such high-multiplicity cluster in the muon endcaps and require the presence of a displaced muon. The most stringent upper limits to date on the branching fraction ℬ( →Φ), where the long-lived particle Φ decays to a pair of hadrons, are obtained for Φ masses of 0.3–3.0 GeV and Φ mean proper decay lengths in the range of 1–500 cm.SCOAP³
Analyzing the Impact of Depth Features on Point Track Performance
Get PDFThe multi-object tracking and segmentation task in urban traffic scenes in for improving autonomous driving, poses ongoing challenges from occlusions, lighting variations, and background noise interferences. We tackle the issue of identity switches by enhancing the existing PointTrack framework, by incorporating raw and monocularly estimated depth information into the color-offset tracking pipeline. By combining depth cues directly into the offset features, our approach strengthens geometric reasoning and leads to improved object association in cases of occlusions and reappearances. On the KITTI multi-object tracking and segmentation dataset, our method reduces identity switching by 21.11% compared to PointTrack baseline, showing increased robustness of tractlet association in challenging scenes. Overall, the approach evaluated notably reduces the occurrence of excessive ID switches, which are a major handicap in real, complicated settings. Numerically, our model performs better by having fewer ID switches while maintaining and in certain cases, enhancing the overall MOTSA score.Ahmet Serhat Yildiz’s Ph.D. is sponsored by the Ministry of National Education of Türkiye
GridNet: A Hybrid LSTM-Based Deep Learning Approach for Accurate Electricity Consumption Forecasting in Smart Grid Systems
Get PDFData Availability Statement:
Upon reasonable request, the corresponding author will provide access to the data that supports
this study.Accurate forecasting of power production and consumption is crucial for optimizing smart grid operations, especially with the growing integration of renewable energy sources, and minimizing CO₂ emissions. This study develops GridNet, a deep learning-based model for forecasting next-hour electricity consumption using the past 24 hours of electricity and time features. GridNet utilizes a two-layer stacked Long Short-Term Memory (LSTM) network with dropout layers to prevent overfitting and improve generalization, followed by a dense output layer. Early stopping and learning rate reduction callbacks were applied to enhance convergence efficiency. The model was optimized using the Adam optimizer and a hybrid loss function combining Mean Squared Error (MSE) with L1 regularization for improved prediction accuracy and robustness. GridNet was trained on a six-year hourly dataset from Romania, covering diverse energy sources like nuclear, wind, solar, for 30 epochs. The performance of the proposed model GridNet was evaluated using several standard metrics, including the Coefficient of Determination (R2), Mean Absolute Error (MAE), MSE, and Root Mean Squared Error (RMSE). Additionally, GridNet's effectiveness was compared with state-of-the-art algorithms, including XGBoost, KNN, Random Forest, and SVM. GridNet outperforms all competing algorithms by up to 59.9% in R2, 72.0% in MAE, 89.8% in MSE, and 68.1% in RMSE, demonstrating its superior accuracy in forecasting electricity consumption trends in smart grids.Kuwait Foundation for the Advancement of Sciences (KFAS)
Advancing Sustainable Agricultural Practices in Africa with AI: Interdisciplinary Approaches to Inclusivity and Resilience
Get PDFArtificial intelligence (AI) is increasingly positioned as a transformative tool in agriculture, yet existing solutions primarily cater to large-scale farms in the Global North, often overlooking the socio-cultural and infrastructural realities of smallholder farmers in Africa. This workshop interrogates how AI can be reimagined to enhance sustainability and resilience in African agriculture by centering farmer agency, cultural knowledge and community and social practice. Building on HCI and CSCW scholarship, we bring together researchers, AI practitioners, NGOs, agronomists, and community stakeholders to explore locally grounded, inclusive, and ethically responsible AI applications. Key themes include trust and skepticism in AI, the role of local languages and epistemologies in model design, strategies for decolonizing AI development and integrating indigenous knowledge and the application of methodologies such as co-design, and participatory AI. The workshop directly aligns with AfriCHI 2025’s theme, "Re-centering African Wisdom in HCI," by fostering interdisciplinary dialogue that embeds African perspectives into AI research and practice. We welcome a variety of contributions including papers, case studies and hands-on demonstrations. Outcomes include a collaboratively developed research agenda, and a white paper synthesizing insights from the workshop. By prioritizing African epistemologies and farmer-centered innovation, this workshop aims to shift AI discourse, ensuring that AI-driven agricultural technologies are not only technically robust but also culturally resonant and socially just
Patenting strategies by pharmaceutical companies: a lawful use of the patent system or an abusive conduct?
Get PDFA preprint of this forthcoming book chapter in Bonadio and Shemtov (ed), ‘A Research Agenda for Patent Law’, Edward Elgar, is available at SSRN: https://ssrn.com/abstract=4591250 or https://doi.org/10.2139/ssrn.4591250. It has not been certified by peer review.To delay generic competition, pharmaceutical companies have been increasingly relying on the patent system. Some of their patent-related practices, such as pay-for-delay agreements, have attracted significant attention of competition authorities in several jurisdictions. However, other practices remain outside of competition authorities’ investigative activities in most jurisdictions, including in Europe. This chapter will discuss two strategies that are employed by pharmaceutical companies to prevent generic competition, namely: (a) strategic accumulation of patents and (b) product hopping. These strategies have the capacity to produce negative effects on the market by delaying market entry of cheaper generics and resulting in a substantial additional cost to the public. The main aim of this chapter is, therefore, to attract attention of policy makers world-wide to these practices explaining their essence, anticompetitive nature and effects
Impact of the Urban Environment on the Thermal Performance and Environmental Quality of Residential Buildings: A Case Study in Athens
Get PDFData Availability Statement:
The original contributions presented in this study are included in the article material. Further inquiries can be directed to the corresponding authors.This paper is an extension of the paper ‘Urban context and climate change impact on the thermal performance and ventilation of residential buildings: a case-study in Athens’ presented at the 43rd AIVC Conference, 4–5 October 2023 Copenhagen, Denmark and the paper ‘Intervention study of climate correlation model predictions for occupant control of indoor environment’ presented at the 44th AIVC Conference, 9–10 October 2024, Dublin, Ireland and published in the conferences’ proceedings.This paper examines the impact of the urban context on the energy performance of a residential building in Athens. Current and future weather files were modified to consider the urban heat island, the overshadowing of adjacent buildings, and the modification of wind speed due to the effects of urban canyons. Dynamic thermal simulations were carried out using the modified weather files. The results indicate that there was a change in heating and cooling demand in comparison to using typical weather files; heating was reduced, but cooling was increased with a total increase in energy demand. There was variation due to height, while overshadowing impacts energy demand significantly. The modified weather analysis also indicates that there are periods in the year that cooling and heating are negligible. During these periods, passive strategies can be used to maintain good internal air quality if occupants are informed how to use their windows and shading devices according to prevailing weather conditions. A method of achieving this occupant-centric operation of the building is described, and the results of an intervention study are discussed. It shows that internal environmental quality can be improved by occupant actions based on forecast weather conditions to direct them.This research was funded by the European Union’s Horizon 2020 research and innovation program under Grant Agreement N° 958345 for the PRELUDE project (https://prelude-project.eu, accessed on 11 March 2025)
Evaluation of Machine Learning and Traditional Statistical Models to Assess the Value of Stroke Genetic Liability for Prediction of Risk of Stroke Within the UK Biobank
Get PDFData Availability Statement:
The data used in this study is available on request from the UK Biobank.Acknowledgments:
This research was conducted using the UK Biobank under Application Number 60549 (www.ukbiobank.ac.uk (accessed on 5 February 2021)). The UK Biobank is generously supported by its founding funders, the Wellcome Trust and the UK Medical Research Council, as well as by the British Heart Foundation, Cancer Research UK, the Department of Health, the Northwest Regional Development Agency, and the Scottish Government. The MEGASTROKE project received funding from sources specified at https://megastroke.org/acknowledgements.html (accessed on 13 September 2022).Supplementary Materials are available online at: https://www.mdpi.com/2227-9032/13/9/1003#app1-healthcare-13-01003 .Background and Objective: Stroke is one of the leading causes of mortality and long-term disability in adults over 18 years of age globally, and its increasing incidence has become a global public health concern. Accurate stroke prediction is highly valuable for early intervention and treatment. There is a scarcity of studies evaluating the prediction value of genetic liability in the prediction of the risk of stroke. Materials and Methods: Our study involved 243,339 participants of European ancestry from the UK Biobank. We created stroke genetic liability using data from MEGASTROKE genome-wide association studies (GWASs). In our study, we built four predictive models with and without stroke genetic liability in the training set, namely a Cox proportional hazard (Coxph) model, gradient boosting model (GBM), decision tree (DT), and random forest (RF), to estimate time-to-event risk for stroke. We then assessed their performances in the testing set. Results: Each unit (standard deviation) increase in genetic liability increases the risk of incident stroke by 7% (HR = 1.07, 95% CI = 1.02, 1.12, p-value = 0.0030). The risk of stroke was greater in the higher genetic liability group, demonstrated by a 14% increased risk (HR = 1.14, 95% CI = 1.02, 1.27, p-value = 0.02) compared with the low genetic liability group. The Coxph model including genetic liability was the best-performing model for stroke prediction achieving an AUC of 69.54 (95% CI = 67.40, 71.68), NRI of 0.202 (95% CI = 0.12, 0.28; p-value = 0.000) and IDI of 1.0 × 10−4 (95% CI = 0.000, 3.0 × 10−4; p-value = 0.13) compared with the Cox model without genetic liability. Conclusions: Incorporating genetic liability in prediction models slightly improved prediction models of stroke beyond conventional risk factors.This research received no external funding
Intelligent Scheduling Methods for Optimisation of Job Shop Scheduling Problems in the Manufacturing Sector: A Systematic Review
Get PDFData Availability Statement:
There are no data associated with this paper.This article aims to review the industrial applications of AI-based intelligent system algorithms in the manufacturing sector to find the latest methods used for sustainability and optimisation. In contrast to previous review articles that broadly summarised existing methods, this paper specifically emphasises the most recent techniques, providing a systematic and structured evaluation of their practical applications within the sector. The primary objective of this study is to review the applications of intelligent system algorithms, including metaheuristics, evolutionary algorithms, and learning-based methods within the manufacturing sector, particularly through the lens of optimisation of workflow in the production lines, specifically Job Shop Scheduling Problems (JSSPs). It critically evaluates various algorithms for solving JSSPs, with a particular focus on Flexible Job Shop Scheduling Problems (FJSPs), a more advanced form of JSSPs. The manufacturing process consists of several intricate operations that must be meticulously planned and scheduled to be executed effectively. In this regard, Production scheduling aims to find the best possible schedule to maximise one or more performance parameters. An integral part of production scheduling is JSSP in both traditional and smart manufacturing; however, this research focuses on this concept in general, which pertains to industrial system scheduling and concerns the aim of maximising operational efficiency by reducing production time and costs. A common feature among research studies on optimisation is the lack of consistent and more effective solution algorithms that minimise time and energy consumption, thus accelerating optimisation with minimal resources.This research received no external funding
Determinants of ThaiMOOC Engagement: A Longitudinal Perspective on Adoption to Continuance
Get PDFData Availability Statement:
The data supporting this study are available upon reasonable request from the corresponding author. Restrictions apply due to a confidentiality agreement established during the ethical approval process at Brunel University London and the conditions set by the gatekeeper who granted permission for data collection. These restrictions specify that the collected data should only be accessed by the researchers involved in this study.Massive Open Online Courses (MOOCs) have become increasingly prevalent in higher education, with the COVID-19 pandemic further accelerating their integration, particularly in developing countries. While MOOCs offered a vital solution for educational continuity during the pandemic, factors influencing students’ sustained engagement with them remain understudied. This longitudinal study examines the factors influencing learners’ sustained engagement with ThaiMOOC, incorporating demographic characteristics, usage log data, and key predictors of adoption and completion. Our research collected primary data from 841 university students who enrolled in ThaiMOOC as a mandatory curriculum component, using online surveys with open-ended questions and post-course usage log analysis. Logistic regression analysis indicates that adoption intention, course content, and perceived effectiveness significantly predict students’ Actual Continued Usage (ACU). Moreover, gender, prior MOOC experience, and specific usage behaviors emerge as influential factors. Content analysis highlights the importance of local language support and the desire for safety during the COVID-19 pandemic. Key elements driving ACU include video design, course content, assessment, and learner-to-learner interaction.This research received no external funding
Data sources:
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! 👇