Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Deep learning for urban air quality: a traffic-based prediction and alert system for Valencia

  • Folgado, Miguel G.
  • Sanz, Verónica
  • Hirn, Johannes
  • Lorenzo-Sáez, Edgar|||0000-0003-3766-1198
  • Urchueguía Schölzel, Javier Fermín|||0000-0002-3054-3431
[EN] Urban traffic congestion is a critical issue with significant implications for mobility and urban planning. In this study, we develop a real-time predictive alarm system capable of forecasting whether a street is likely to experience unusually high traffic within the next 30 min. The system classifies road segments into three alert levels based on traffic data updated every 10 min, providing timely information that can support decision-making in traffic management. The prediction model is built using deep learning techniques trained on a whole year of traffic data in the city of Valencia, and tested with the following year¿s data. We evaluated different neural network architectures, including long short-term memory (LSTM) networks, an extended LSTM variant (xLSTM), and Graph Neural Networks (GNNs). Our results show that LSTM provides the best balance between accuracy and computational efficiency, making it the most suitable model for real-time deployment. In addition to traffic data, we incorporate meteorological variables such as wind speed, wind direction, and precipitation to explore their potential impact on traffic dynamics. Although the relationship between traffic and environmental conditions warrants further study, this work demonstrates the feasibility of using real-time predictions to improve urban mobility strategies. The proposed system offers a data-driven approach that can be integrated into broader traffic management frameworks to improve efficiency and responsiveness., The research of MGF is supported by the Margarita Salas postdoctoral fellowship from the Ministerio de Educación MS21-038 and the Generalitat Valenciana ASFAE/2022/020. The research of VS is supported by the Generalitat Valenciana PROMETEO/2021/083 and the Ministerio de Ciencia e Innovacion PID2020-113644GB-I00. The research of JH is supported by the RESECARIN project TED2021-129682B-I00 from the Spanish Ministerio de Ciencia e Innovación. The research of JFU and ELS is supported by the Agència Valenciana de la Innovació (AVI) INNEST/2021/263.
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Evaluating the restoration of plant ecological interactions in gypsum mines with species co-occurrences analyses

  • Hirn, Johannes
  • Sanz, Verónica
  • Lorite, Juan
  • Martínez-Hernández, Fabián
  • Mendoza-Fernández, Antonio Jesús
  • Mota, Juan Francisco
  • Navarro-Cano, J. A.
  • Pérez-García, Francisco Javier
  • Prieto-Rubio, Jorge
  • Sánchez‑Martín, Ricardo
  • Verdú, Miguel
Ecological interactions are a crucial component of biodiversity, and their loss can disrupt ecosystem functions. Therefore, restoring these interactions is essential for effective ecological restoration. Although positive interactions between plants (i.e. plant facilitation) have gained attention in restoration ecology, most research has concentrated on population and community outcomes, with limited focus on the restoration of plant–plant interactions. We used Markov network models to evaluate pairwise and higher-order plant interactions in gypsum shrublands in southern Spain and compared them with interactions in adjacent areas restored after mining activities. Restoration efforts, involving seed-based and seedling-based approaches from 13 years ago, were assessed to determine whether positive (facilitation) and negative (competition) interactions had been restored within each community. Given the stressful conditions of gypsum soils, we hypothesize that facilitation is the main force shaping the interactions in both reference and restored areas. Positive interactions between the most abundant species dominated over negative ones in both natural and restored areas. In both the seed-based and the seedling-based restorations, about half of the observed pairwise interactions in the natural area were restored with the same sign. Generally, facilitation interactions that were not restored became neutral, but rarely turned negative. Synthesis and applications. These case studies suggest that both seed-based and seedling-based approaches can restore pairwise species interactions to a similar extent. The seedling-based method, although more expensive, provides immediate visual impact by directly introducing canopy-forming species into the system. In contrast, the seed-based approach, which is more cost-effective, can establish highly dense plant communities, enhancing soil protection, though it may also lead to increased emergent competition through third-order interactions., Financial support was provided by the projects TED2021-129682B-I00, PID2020-113157GB-I00, PID2023-146535NB-I00 (funded by MCIN/AEI/10.130 39/501100011033 and ‘ERDF A way of making Europe’) and CIPROM/2021/63 (Generalitat Valenciana). Alba Navarro, Daniel A. Rodríguez, and Santiago Donat helped with field work., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

Reimagining plant science training in the era of generative artificial intelligence: a global perspective

  • Moghe, Gaurav D.
  • Zimić-Sheen, Alen
  • Chen, Dijun
  • Yadav, Gitanjali
  • Cao, Guangshuo
  • Tufan, Hale
  • Williams, Jason
  • Szymański, Jędrzej
  • Kim, Jeongwoon
  • Busta, Lucas
  • Mutwil, Marek
  • Verdú, Miguel
  • Zimić, Mirko
  • Provart, Nicholas J.
  • Makunga, Nokwanda
  • Wilkins, Olivia
  • Sun, Qi
  • VanBuren, Robert
  • Marks, Rose A.
  • Rhee, Seung Y.
  • Jiang, Yu
  • Xie, Yuying
In recent years, a deluge of big and diverse datasets from hundreds of plant species, coupled with spectacular innovations in artificial intelligence (AI) and generative AI (GenAI), has altered the landscape of plant science. These developments are increasingly democratizing the field, reducing the entry barriers to complex data analysis and enabling a new wave of innovative research while introducing new challenges. Therefore, in this era, it is critical that we train the next generation of plant scientists to be AI-literate, ie, not only proficient in using AI but also vigilant about its pitfalls and biases. In this perspective, we call for six strategic shifts necessary for training the next generation of plant scientists. We argue that while maintaining a core focus on subject expertise, educators should simultaneously emphasize development of new AI-forward pedagogical and evaluation frameworks that reward interdisciplinary and critical thinking, human-driven knowledge synthesis, self-directed learning, and conceptual understanding of workflows. For effective critique and sound interpretations based on biological reality, plant scientists must be explicitly trained in recognizing biases underlying GenAI models. Finally, we highlight the structural barriers hindering the equitable and ethical use of GenAI, where awareness and resolution are critical for sustainable growth of the field. Through the above conceptual framework and numerous plant-science-focused illustrative activities, examples, and resources meant for students and educators alike, this Perspective defines high-level emphasis areas for GenAI-enabled scientific training, aimed at creating a more effective, engaged, and adaptive community of plant scientists., This work was supported by U.S. National Science Foundation awards OISE-2434687 and IOS-2310395 to Moghe (G.M.), Ministerio de Ciencia e Innovación TED2021-129682B-I00 to Verdu (M.V.), Research Corporation For Science Advancement CS-CSA-2025-040 to Busta (L.B.), Novo Nordisk Starting Grant to Mutwil (M.M.), US NSF awards OISE-2434687, IOS-2312181, IOS-2406533, MCB-2420360, DBI-2419923 and U.S. Department of Energy awards DE-SC0018277, DE-SC0020366, DE-SC0023160, DE-SC0021286, DE-SC0008769 to Rhee (S.R.), Natural Sciences and Engineering Research Council of Canada Global Alliance Discovery Grants Program Grant ALLRP 597259-24 to Provart (N.P.), National Research Foundation (South Africa) award CPRR230503101428 to Makunga (N.M.), The National Natural Science Foundation of China award T2541063 to Chen (D.J.), Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy award EXC-2048/1—project ID 390686111 to Szymański (J.S.)., Peer reviewed