Federated Learning of Oligonucleotide Drug Molecule Thermodynamics with Differentially Private ADMM-Based SVM
Paper i proceeding, 2021

A crucial step to assure drug safety is predicting off-target binding. For oligonucleotide drugs this requires learning the relevant thermodynamics from often large-scale data distributed across different organisations. This process will respect data privacy if distributed and private learning under limited and private communication between local nodes is used. We propose an ADMM-based SVM with differential privacy for this purpose. We empirically show that this approach achieves accuracy comparable to the non-private one, i.e. ∼ 86 %, while yielding tight empirical privacy guarantees even after convergence.

Federated learning

Oligonucleotide drug molecules

SVM

ADMM

Differential privacy

Distributed learning

Författare

Shirin Tavara

Data Science och AI 1

Alexander Schliep

Göteborgs universitet

Debabrota Basu

Université de Lille

Communications in Computer and Information Science

1865-0929 (ISSN) 18650937 (eISSN)

Vol. 1525 CCIS 459-467
9783030937324 (ISBN)

21st European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2021
Virtual, Online, ,

Ämneskategorier

Bioinformatik (beräkningsbiologi)

Bioinformatik och systembiologi

Datavetenskap (datalogi)

DOI

10.1007/978-3-030-93733-1_34

Mer information

Senast uppdaterat

2022-03-24