Luca Carlone’s Post

🔗 Check out an exciting new blog post on distinguishing neighborhood representations through the reverse process of GNNs for heterophilic graphs. The post delves into applying the reverse process design principle to three variants of GNNs and its impact on prediction performance in heterophilic graph data. Discover how the reverse mechanism can mitigate oversmoothing and improve classification. Read more at https://bit.ly/49UST3R 📚 #SocialMediaMarketing #GNN #GraphNeuralNetworks #DataScience

I am happy to share the recent achievement of my PhD. student, #Shadi Bikas, whose work has been accepted by IEEE Transactions on Network and Service Management (TNSM). In our work, we showed the impact of MPTCP paths on Quality of Experience (QoE). Curious about our findings? Please check out the full paper: https://lnkd.in/e6Jyich7

Check out my latest article for The Association of Reproductive and Clinical Scientists , where I reflect on my first year as an STP trainee! 🔬🥼🎉 It has been an amazing year, and I cannot wait for what is still to come! Follow the links below to read my article, and others, as part of the Autumn edition of The Reproductive Scientist!

In this Outlook, Wright and Wilusz discuss recent findings by Xiao et al. that show that lncRNA Malat1 is exported to the cytoplasm and translated in cultured neurons. Read the Outlook here: ➡️ https://lnkd.in/eKYAEqys For the full story: ➡️ https://lnkd.in/eK8z5rC3

Center for Targeted Machine Learning and Causal Inference (CTML) Faculty Laura B. Balzer and Biostatistics Alumni Alejandra Benitez are presenting at the WNAR/IMS Conference on "Clustering, Weighting, & Estimating: Lessons from the Preterm Birth Initiative." This conference occurs in Fort Collins, Colorado from June 9 to June 12. Learn more about their talk below! This talk reviews traditional and novel algorithms to estimate effects in cluster randomized trials (CRTs), comparing their practical performance, overall performance, and when there are few clusters of different sizes. Utilizing the Preterm Birth Initiative (PTBi) study, the presenters demonstrate their recommendation of TMLE for CRT analysis and real-world impact. #ctml #berkeleyCTML #ucberkeley #WNAR2024

New paper out today, based on the MSc work of Tore von Schwartz. We show how to use representation learning to organize different models in latent space. This allows us to distinguish between different models, select benchmark scenarios to be searched for at the LHC, and identify possible gaps in model space. https://lnkd.in/dmwpB8hz

NOTABLE ARTICLE, September 2024 #MedicalPhysics: "Improved brain metastases segmentation using generative adversarial network and conditional random field optimization mask R‐CNN." #MPJ #SeptemberNotableArticle Please join us in congratulating Haowen Pang and their team whose article was selected as a notable paper. https://lnkd.in/gT3Hi9Xz

ICYMI: Ana Maria Rey gave a talk at the #MCQSTColloquium last month on "Optical Lattice Clocks: From Timekeepers to Spies of the Quantum Realm" The recorded talk is available on our YouTube channel. Watch it now and discover other insightful talks into #QuantumScience and #QuantumTechnology. 🎥 https://lnkd.in/d_xfxmdF

Symposium "AMR Forensics" - Global tracking of antimicrobial resistance genes Program and speakers: https://amrforensics.org/ Date: Monday 8 April Location: Utrecht Free of charge, but please register: https://amrforensics.org/ It remains an open question where and how antimicrobial resistance (AMR) genes migrate across the globe. While targeted monitoring by sequencing is an important tool to address this question, we are convinced that no one party can do this alone. To meaningfully combine data from different sources, including the wealth of microbial sequence data that is already available in public repositories, requires innovative high-performance bioinformatics tools and high-quality metadata. This could broaden the scope of AMR forensics, streamline the process of uncovering global trends, and provide valuable insights for policymakers This symposium is co-organized by Utrecht University and Koninklijke Nederlandse Vereniging voor Microbiologie (KNVM) Section Microbial Bioinformatics, and supported by ZonMw

Magnetoencephalographic signatures of hierarchical rule learning in newborns. Results suggest that infants in the first weeks of life are able to learn hierarchical rules, yet a certain level of vigilance seems to be necessary. https://lnkd.in/gC_a6veR