Procedia Computer Science最新文献

{"title":"Towards full AI model lifecycle management on EuroHPC systems, experiences with AIFS for DestinE","authors":"Thomas Geenen ,&nbsp;Even Marius Nordhagen ,&nbsp;Victor Sanchez ,&nbsp;Cathal O'Brien ,&nbsp;Simon Lang ,&nbsp;Mihai Alexe ,&nbsp;Ana Prieto Nemesio ,&nbsp;Gert Mertes ,&nbsp;Rakesh Prithiviraj ,&nbsp;Jesper Dramsch ,&nbsp;Baudouin Raoult ,&nbsp;Florian Pinault ,&nbsp;Helen Theissen ,&nbsp;Sara Hahner ,&nbsp;Mario Santa Cruz ,&nbsp;Matthew Chantry ,&nbsp;Nils Wedi","doi":"10.1016/j.procs.2025.02.264","DOIUrl":"10.1016/j.procs.2025.02.264","url":null,"abstract":"<div><div>On October 13 2023 ECMWF released the first alpha version of its artificial intelligence forecasting system, AIFS, ECMWFs data-driven forecasts model. This first release came just a few months after ECMWF started the development of this new model that highlights the increased efforts in the field of machine learning (ML) that ECMWF has been building over the last few years. This paper describes the use of AIFS on EuroHPC systems in the context of DestinE. The main focus is on performance benchmarks on the different EuroHPC systems available to DestinE but also very much on the deployment and use of the tools to support the model lifecycle management. EuroHPC systems have already proven to be of great value for DestinE and in this paper, we describe how we leverage these systems for artificial intelligence (AI) and ML models in DestinE. We are closely working with EuroHPC and EuroHPC hosting sites through co-design and the optimization of existing solutions to optimize the usage of these systems in every step of the lifecycle management for AI and ML models. The performance benchmarks of our models on several EuroHPC systems showed that the speedup is close to linear up to several thousand GPUs, but that for each EuroHPC system a different optimization strategy must be used to achieve that. For model lifecycle management we found that we can use our in-house developed, domain specific, framework on EuroHPC systems and highlight some specific modifications and future improvements for EuroHPC systems. W e a l s o provide implementation details and share our experiences on how to retrieve and collect provenance data and information from models running on EuroHPC systems using (external to the EuroHPC system deployed) cloud native frameworks. Although we describe solutions in this paper that are designed to support our specific requirements and context, we believe that proposed solutions, developments and implementation details can also bring value beyond the broader NWP community.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"255 ","pages":"Pages 93-102"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143563317","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Portable test run of ESPResSo on EuroHPC systems via EESSI","authors":"Alan O'Cais ,&nbsp;Kenneth Hoste ,&nbsp;Jean-Noël Grad ,&nbsp;Caspar van Leeuwen ,&nbsp;Lara Peeters ,&nbsp;Satish Kamath ,&nbsp;Thomas Röblitz ,&nbsp;Richard Topouchian ,&nbsp;Bob Dröge ,&nbsp;Pedro Santos Neves ,&nbsp;Rudolf Weeber","doi":"10.1016/j.procs.2025.02.267","DOIUrl":"10.1016/j.procs.2025.02.267","url":null,"abstract":"<div><div>One of the milestones of the EuroHPC Centre of Excellence MultiXscale is to be able to run the EESSI test suite on at least two different architectures available on EuroHPC Supercomputers. Our initial efforts focused on making the test suite portable across two different supercomputers: Karolina and Vega (the CPU partitions of both are a Zen2 micro-architecture).</div><div>More recently we have spent time getting the same test suite working on a more “exotic” architecture, the ARM A64FX architecture of Deucalion (which was in pre-production at the time of the experiment). This has raised some additional complications for EESSI as CernVM-FS (which is used to distribute EESSI) was not yet natively available there.</div><div>We show the current scalability of the ESPResSo application using the portable test suite. ESPResSo is already known to have scalability issues for both multi-node and multi-GPU configurations which are currently being analysed in collaboration with the POP Centre of Excellence. The purpose of this effort was to ensure that we can quickly and automatically record the performance of the application across a range of EuroHPC systems (i.e. ESPResSo acts as a pilot application for the full test suite)</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"255 ","pages":"Pages 122-129"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143563320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Tabletop exercises as a potential to improve \"Preparedness for Society in Health Crises and Disasters\"","authors":"Felix Baumann ,&nbsp;Lennart Landsberg ,&nbsp;Thomas Säger ,&nbsp;Ompe Aimé Mudimu","doi":"10.1016/j.procs.2025.02.101","DOIUrl":"10.1016/j.procs.2025.02.101","url":null,"abstract":"<div><div>The PREPSHIELD project, acronym for \"Preparedness for Society in Health Crises and Disasters\", is part of the \"HORIZON-CL3-2023-DRS-01-01\" funding program titled \"Improving Social and Societal Preparedness for Disaster Response and Health Emergencies\". It aims to enhance preparedness for health crises by incorporating both social and societal elements, highlighted by the urgent needs revealed during the COVID-19 pandemic. Focusing on a people-centered approach, PREPSHIELD leverages AI technology to develop strategies and platforms that address the social impacts of health emergencies. Spearheaded by the Università degli Studi del Piemonte Orientale with a €3,891,762.5 budget, this initiative unites a diverse group of stakeholders, including government, civil society, and healthcare entities, to foster tailored communication, management, and simulation tools for anticipating future crises. The project features drills and simulations across local to national scales, supported by a digital platform leveraging the CRIMSON model for effective crisis communication. Bringing together five universities, numerous government and non-profit organizations, an SME, and large companies from seven EU member states and Switzerland, PREPSHIELD seeks to significantly advance preparedness and response capabilities for health crises through innovative technology and a comprehensive, collaborative approach. The results of the project are expected to be methods and approaches that provide evaluation tools and metrics for reviewing crisis management. In addition, proposals for action will be developed for the context of crisis management in health crises.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"256 ","pages":"Pages 101-105"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143592856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A systematic review of active learning approaches in the selection of medical images","authors":"Maria Santos ,&nbsp;Goreti Marreiros","doi":"10.1016/j.procs.2025.02.186","DOIUrl":"10.1016/j.procs.2025.02.186","url":null,"abstract":"<div><div>Background: Active Learning has been proven to be an effective way to maximize the model’s learning capacity, using fewer amounts of labeled data. In the field of medical imaging data, data and annotations can be scarce and very expensive to obtain, so techniques like Active Learning can be a useful solution. Methods: For this systematic review, the data sources were obtained through IEEE Explore, PubMed, and ACM Digital Library, between the period of 2018 and 2023. Only studies that belonged to the field of healthcare (using medical images as a dataset) and machine learning, written in English and that were not a book, or a survey were used. Covidence was used as a tool to synthesize the results. Results: From 336 records, 51 were included in this review. Interpretation: Most studies showed that Active Learning can have a positive impact on the construction of models, however, it is important to not consider only the informativeness/uncertainty of the sample, but also the distribution of the data, reducing the probability of selecting samples that are not representative enough of the dataset or outliers. Active Learning is usually an iterative process until a stop criterion is met, for example, the model’s performance. To evaluate an Active Learning solution, the proposed method is usually compared with random sampling, or other Active Learning queries.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"256 ","pages":"Pages 843-851"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143592862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"K-operator as a predictor for Alzheimer-Perusini’s disease","authors":"Maria Mannone ,&nbsp;Norbert Marwan ,&nbsp;Peppino Fazio ,&nbsp;Patrizia Ribino","doi":"10.1016/j.procs.2025.02.173","DOIUrl":"10.1016/j.procs.2025.02.173","url":null,"abstract":"<div><div>Progressive memory loss occurring in age-related neurological diseases contributes to the disgregation of the individual, with serious personal and social consequences. We model the brain network damage provoked by a neurological disease through a physics-inspired mathematical operator, <em>K</em>. Acting on a diseased brain, <em>K</em> provides the disease time evolution. Focusing on Alzheimer-Perusini’s disease (AD), we approximate the <em>K</em>-operator considering selected patients of the ADNI 2 dataset. We also propose <em>K</em> as a predictor for the disease progress over time and give its preliminary evaluation in the AD progression from the cognitive normal (CN) stage to AD through intermediate mild cognitive impairment (MCI) stages.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"256 ","pages":"Pages 731-738"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143593269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Neural Network Ensemble for Detecting Parasite Eggs in Microscopic Images","authors":"Matheus L.L. Bessa ,&nbsp;Geraldo Braz Junior ,&nbsp;João Dallyson Souza de Almeida","doi":"10.1016/j.procs.2025.02.174","DOIUrl":"10.1016/j.procs.2025.02.174","url":null,"abstract":"<div><div>Intestinal parasite infections are a global health problem. In 2022, the WHO estimates that up to 1.2 billion people will be infected with Ascaris lumbricoides. Diagnosis is conducted by analyzing faecall samples under a microscope. However, this process is laborious and prone to error. Considering this, this study proposes a methodology to automate the detection of parasite eggs in microscope images. This methodology applies multiple object detectors in an ensemble and submits a model to reduce false negatives in the public dataset Chula-ParasiteEgg-11, with 11,000 images and 11 classes of parasites. Using this approach, it was possible to reduce the false negative rate and improve the f1 score up to 0.94. The results suggest that the proposed model leads to a reduction of false negatives and an improvement in recall.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"256 ","pages":"Pages 739-746"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143593270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Strategies for Improved Out-of-Distribution Detection in Drone vs. Bird Classification","authors":"Ami Pandat ,&nbsp;Punna Rajasekhar ,&nbsp;Gopika Vinod ,&nbsp;Rohit Shukla","doi":"10.1016/j.procs.2025.03.341","DOIUrl":"10.1016/j.procs.2025.03.341","url":null,"abstract":"<div><div>The use of drones has expanded significantly across various applications over the past decade, leading to increased surveillance-related challenges. These challenges raised the necessity of developing Anti-Drone systems. One of the critical requirements for an effective Anti-Drone system is the ability to accurately distinguish drones from birds in the sky. While deep learning-based classification techniques have been employed for this task, they often suffer from high false positive rates. To address this challenge, Out-of-Distribution (OOD) detection is essential for enhancing the reliability and robustness of drone surveillance systems, particularly in differentiating drones from birds. This paper explores several techniques to improve OOD detection performance, focusing on Energy-Based Models (EBM) and Variational Autoencoders (VAE). We evaluate four loss functions within the EBM framework: Mean Squared Error (MSE) Loss, Mean Squared Error with OOD Penalty, Contrastive Loss, and Binary Cross-Entropy with Energy Regularization. Our results demonstrate that the Mean Squared Error with OOD Penalty function achieves the best performance, with an AUC of 0.9, providing clearer separation between in-distribution (drones) and out-of-distribution (birds) samples. However, the VAE approach did not yield significant results for the binary classification task. Future work could explore hybrid approaches to further enhance OOD detection in such applications.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"259 ","pages":"Pages 398-407"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143936370","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transformative Tech and Social Dynamics: Redefining Customer Engagement in Industry 5.0","authors":"Bhawana Kothari ,&nbsp;Ambica Prakash Mani ,&nbsp;V M Tripathi","doi":"10.1016/j.procs.2025.03.325","DOIUrl":"10.1016/j.procs.2025.03.325","url":null,"abstract":"<div><div>In the era of Industry 5.0, this paper offers a thorough assessment of how transforming technologies have changed consumer involvement. We start by looking at the change from conventional, linear methods to client participation towards current, technologically driven techniques. The emphasis is on how digitization is drastically changing consumer expectations and the policies companies have to follow to quickly fit these developments. Examining the ability of modern technologies such as artificial intelligence (AI), machine learning, the Internet of Things (IoT), and blockchain to influence modern customer engagement methods takes up a good amount of this research. We look at how machine learning and artificial intelligence improve predictive analytics, therefore allowing companies to proactively satisfy consumer wants and customize experiences. We also explore the key part IoT plays in creating a consistent and flawless consumer experience at many points of contact. This talk also addresses the creative use of virtual assistants and chatbots, assessing their efficiency in providing real-time consumer help. These AI-driven solutions are evaluated for their capacity to provide a degree of customizing like that of human interactions, hence improving client connections. This paper explores several aspects of customer commitment in the computerized era using approaches, innovations, and best practices that enable companies to create further associations, enhance customer interactions, and propel supportable development in an era marked by mechanical disturbance.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"259 ","pages":"Pages 240-249"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143936372","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent Analysis Method of E-commerce Data Based on Various Machine Learning Algorithms","authors":"Bo Yang","doi":"10.1016/j.procs.2025.04.008","DOIUrl":"10.1016/j.procs.2025.04.008","url":null,"abstract":"<div><div>Under the current rapid development of the e-commerce industry, most e-commerce companies are pursuing to enhance the clicks of products and its conversion rate to buy. And there are many machine learning algorithms for the intelligent analysis of e-commerce data, among which, the most widely used is the recurrent neural network (RNN) and collaborative filtering algorithm. Based on the use of multiple machine learning algorithms, this paper compares the differences in the clicks of products and the purchase conversion rates between the RNN algorithm and the collaborative filtering algorithm. The RNN algorithm can make full use of the behavior sequence time dependence and context information and the collaborative filtering algorithm is based on the similarities between user and product. The evaluation results are as follows: the products clicked by the RNN algorithm are between 18,000 and 25,000, which is significantly higher than the products clicked by the collaborative filtering algorithm. In order to improve user purchase decisions and overall sales efficiency, e-commerce operators can select the RNN algorithm to fully understand the user’s interests and needs, and provide accurate personalized product recommendations.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"259 ","pages":"Pages 591-597"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143936924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Advanced Video-Based Deep Learning Framework for Comprehensive Detection, Diagnosis, and Classification of Dermatological Conditions in Real-Time Datasets","authors":"Syed Thouheed Ahmed ,&nbsp;Amogh S Guthur ,&nbsp;Pratyush Kumar Rai ,&nbsp;Pranava Swaroop N","doi":"10.1016/j.procs.2025.03.344","DOIUrl":"10.1016/j.procs.2025.03.344","url":null,"abstract":"<div><div>The advanced acne detection model showcased in this project utilizes deep learning methods to accurately classify skin conditions, including blackheads, dark areas, blemishes, and creases. It employs a YOLOv5 format annotation scheme to analyze spatial and temporal information from video sequences, resulting in exceptional performance in detecting seven distinct classes. The model’s resilient performance indicates high accuracy, with a mean Average Precision (mAP) of about 0.85-0.9 at an IoU threshold of 0.5. It also demonstrates generalization and robustness with an mAP of 0.5-0.55 across IoU thresholds from 0.5 to 0.95, making it suitable for real-world dermatological assessments. The proposed method enables early detection and more effective treatment by monitoring skin conditions over time, significantly impacting dermatological image analysis. The goal is to improve patient outcomes and provide personalized skincare recommendations using deep learning techniques, benefiting clinicians and researchers in analyzing and categorizing skin conditions. Additionally, incorporating additional data sources like clinical images or medical histories can enhance the model’s diagnostic capabilities and accuracy. Expanding the dataset will enhance the model’s generalizability and robustness for new skin conditions</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"259 ","pages":"Pages 424-432"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143936476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}