3 papers
cs.LG2026
DP-Hype: Federated Differentially Private Hyperparameter Search
Johannes Liebenow, Thorsten Peinemann, Esfandiar Mohammadi
Tuning hyperparameters in federated machine learning can substantially impact model performance. When hyperparameters are tuned on sensitive data, privacy becomes an important chal…
cs.LG2026
Non-omniscient backdoor injection with one poison sample: Proving the one-poison hypothesis for linear regression, linear classification, and 2-layer ReLU neural networks
Thorsten Peinemann, Paula Arnold, Sebastian Berndt +2
Backdoor poisoning attacks are a threat to machine learning models trained on large data collected from untrusted sources; these attacks enable attackers to inject malicious behavi…
cs.CR2024
S-BDT: Distributed Differentially Private Boosted Decision Trees
Thorsten Peinemann, Moritz Kirschte, Joshua Stock +2
We introduce S-BDT: a novel -differentially private distributed gradient boosted decision tree (GBDT) learner that improves the protection of single training data…