Logo des Repositoriums

Auflistung nach Schlagwort "Privacy Aware Machine Learning"

1 - 2 von 2
  • Textdokument
    Learning without Looking: Similarity Preserving Hashing and Its Potential for Machine Learning in Privacy Critical Domains
    (INFORMATIK 2022, 2022) Eleks,Marian; Rebstadt,Jonas; Fukas,Philipp; Thomas,Oliver
    Machine Learning is frequently ranked as one of the most promising technologies in several application domains but falls short when the data necessary for training is privacy-sensitive and can thus not be used. We address this problem by extending the field of Privacy Aware Machine Learning with the application of Similarity Preserving Hashing algorithms to the task of data anonymization in a Design Science Research approach. In this endeavor, novel anonymization algorithms made to enable Machine Learning on anonymized data are designed, implemented, and evaluated. Throughout the Design Science Research process, we present a collection of issues and requirements for Privacy Aware Machine Learning algorithms along with three Similarity Preserving Hashing-based algorithms to fulfil them. A metric-based comparison of established and novel algorithms as well as new arising opportunities for Machine Learning on sensitive data are also added to the current knowledge base of Information Systems research.
  • Konferenzbeitrag
    Privacy Aware Processing
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Eleks, Marian; Rebstadt, Jonas; Kortum, Henrik; Thomas, Oliver
    In many machine learning (ML) applications, the provision of data and the training as well as the analysis of machine learning systems are performed by distinct actors, a data owner and a data consumer. To protect sensitive information in these ML-scenarios, privacy aware machine learning (PAML) methods are often applied to the data before sharing. Based on the type of PAML methods used, data understanding and preparation as defined in the CRISP-DM model become more difficult if not impossible. To enable these steps, we propose a method to share a variety of uncritical information with the data consumer who is then able to define the necessary processing steps on a meta-level. These are then applied to the data in the data owners local trusted environment before the PAML-methods whereupon the prepared and protected data is shared.