Rudolf Mayer

[FeatureCloud @Feature_Cloud]

📚Publication alert! Our main #FeatureCloud consortium paper is out, describing in detail the structure and versatility of the #AppStoreHealthcare and how it simplifies the development and execution of #FederatedLearning algorithms. https://t.co/TT2PU7k9r6
@HorizonEU @cosybio_UHH

[FeatureCloud @Feature_Cloud]

🔐Privacy in Medicine: K-anonymity is an approach for #privacy-preserving publishing of personal, sensitive data. As a result of the #anonymization, however, the data utility may suffer. Thus, quantifying #UtilityLoss is important! https://t.co/tGJBfrKAHw @SBA_Research @HorizonEU

[FeatureCloud @Feature_Cloud]

🚨How can #anomalies such as fraud, intrusions, or rare diseases be identified in #FederatedLearning? This paper studies anomaly detection on two #datasets in supervised, semi-supervised, and unsupervised settings vs. a centralized approach. https://t.co/WzY7cKUQvg @SBA_Research

[FeatureCloud @Feature_Cloud]

🔐Privacy in #FederatedLearning & #AI: Can an honest-but-curious participant determine whether a data sample was used in the training process of a machine learning model? Does this pose a risk to #privacy? Find out in this paper: https://t.co/rBoH39FsRF @SBA_Research @HorizonEU

[FeatureCloud @Feature_Cloud]

🔐Cyber Risks: This paper investigates #attack scenarios & success rates for a #malicious node in #FederatedLearning, considering both sequential and parallel strategies, as a basis for estimating risks from potential adversaries. https://t.co/5FEgi134gb @SBA_Research @HorizonEU

[FeatureCloud @Feature_Cloud]

🔐Cyber Risks in #FederatedLearning: By altering inputs in the #model #training phase, an adversary may later trigger #malicious behavior in the prediction phase. This paper describes scenarios similar to traffic sign & face recognition data. @SBA_Research https://t.co/IeONQQjHb8

[FeatureCloud @Feature_Cloud]

Meet the #FeatureCloud team: The passion and expertise of @SBA_Research (Vienna, Austria) is to enhance #CyberSecurity.

@Rudolf_Mayer leads WP2 - "Cyber risk assessment and mitigation". @walidfdhila leads WP6 - "Blockchains and user rights management".
https://t.co/uEkfNYEp4m

[FeatureCloud @Feature_Cloud]

Meet the #team: @SBA_Research members working on #FeatureCloud also include PhD students Anastasia Pustozerova & @SarcevicTanja, junior researchers Daryna Oliynyk & Daniela Martinez Duarte, and senior researcher Aljosha Judmayer. Thanks for your hard work!
https://t.co/uEkfNYEp4m

[FeatureCloud @Feature_Cloud]

🔍This new #FeatureCloud review summarizes the current discussion on legal concerns & #GDPR compliance related to #FederatedLearning systems, including #privacy-enhancing technologies, e.g. differential privacy & secure multiparty computation in medicine. https://t.co/PEUC7QVr2C

[FeatureCloud @Feature_Cloud]

🖥️This paper highlights #FederatedLearning as a 🔐#privacy-aware data mining strategy and investigates the #DataLeakage of 3 popular algorithms for so-called "QR decomposition" (Gram-Schmidt orthonormalization, Householder algorithm, and Givens rotation). https://t.co/UrhqHHwTQN

[FeatureCloud @Feature_Cloud]

💡#IP protection in #MachineLearning (ML): Commercial use of #ML is spreading while IP protection of trained #models remains an issue. This article uses a comprehensive threat model, categorizes attacks & defenses, and consolidates taxonomy. @SBA_Research https://t.co/eqkFCwTTi2

[FeatureCloud @Feature_Cloud]

🧬#SystemsMedicine today⚕️: This #Review discusses #MolecularNetworks (types of data, analytical methods, ways to validate them). Categorizing diseases by organs or symptoms is the past - deciphering disease mechanisms is the future! @REPO4EU @hhhwschmidt
https://t.co/pyjYsBlCG7

[FeatureCloud @Feature_Cloud]

At this year's German Conference on Bioinformatics @gcb2023, #FeatureCloud organized a productive workshop on #privacy-aware #FederatedLearning in biomedicine. Thank you for attending & contributing ideas to make the platform stronger! @janbaumbach @cosybio_UHH @unihh @HorizonEU

[REPO4EU @REPO4EU]

Meet the #RExPO23 speakers 👇

🔵Prof @hhhwschmidt from @MaastrichtU

🔵Paolo Parini from @karolinskainst & @KarolinskaUnsju

🔵Annika Tibell from @KarolinskaUnsju

🔵Robert Harris from @karolinskainst

Check the full agenda and list of speakers ⤵️
https://t.co/Ya250cgoen

[gcb2025 @gcb2024]

Don't be sad that #GCB2023 is over - save the date and look forward to #GCB2024! See you in Bielefeld from 30 Sep - 2 Oct 2024 @DECHEMA @unibielefeld @BIBIBielefeld @CeBiTec @denbiOffice @fz_juelich @cosybio_UHH

[FeatureCloud @Feature_Cloud]

It's ready 📽: https://t.co/1owecAdWoZ
Watch the summary clip about our 10th FeatureCloud Steering Committee (SC) Meeting in Romania🐻, 11-13 June 2023. #FeatureCloud #AppStoreHealthcare #H2020 @janbaumbach @cosybio_UHH @MedUni Graz@SBA_Research @researchinst @Heiderlab @imadaSDU

[FeatureCloud @Feature_Cloud]

A lot is happening at the intersection of next-generation #sequencing🧬and #AI 💻, especially in #CancerResearch: This #FeatureCloud review paper provides in-depth insight into this fascinating field! @Heiderlab @ypark_bioinfo @achauschild @MDPIOpenAccess https://t.co/izDTpW7xka

[FeatureCloud @Feature_Cloud]

#BigData analysis🧬faces many obstacles. For example, disease module mining methods (DMMMs) often include non-robust steps in their workflows. To overcome this issue, we here present "ROBUST", a new DMMM! #robustness #scalability #omics #H2020 @HorizonEU
https://t.co/mSfEhceLae

[FeatureCloud @Feature_Cloud]

This collaborative work in @Nature_NPJ used #InVitro models to unravel regulatory functions of #oncogene CTCFL 🧬- a transcriptional factor expressed in #OvarianCancer. Candidate genes were assessed regarding prognostic potential & druggability? #H2020 https://t.co/7DBMrZIAwf

[CoSy.Bio @cosybio_UHH]

The Steering Committee of the Horizon Europe project @MicrobAIome_EU, coordinated by @janbaumbach, met this week in France! We are excited for the potential of the project to help lower mortality due to #ColorectalCancer using the power of #AI and #Microbiomes!