On reconstructability of quadratic utility functions from the iterations in gradient methods
2016 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 66, 254-261 p.Article in journal (Refereed) PublishedText
In this paper, we consider a scenario where an eavesdropper can read the content of messages transmitted over a network. The nodes in the network are running a gradient algorithm to optimize a quadratic utility function where such a utility optimization is a part of a decision making process by an administrator. We are interested in understanding the conditions under which the eavesdropper can reconstruct the utility function or a scaled version of it and, as a result, gain insight into the decision-making process. We establish that if the parameter of the gradient algorithm, i.e., the step size, is chosen appropriately, the task of reconstruction becomes practically impossible for a class of Bayesian filters with uniform priors. We establish what step-size rules should be employed to ensure this.
Place, publisher, year, edition, pages
Elsevier, 2016. Vol. 66, 254-261 p.
Statistical inference, Data privacy, Gradient methods, Data confidentiality, Parameter identification, Quadratic programming
IdentifiersURN: urn:nbn:se:kth:diva-184949DOI: 10.1016/j.automatica.2016.01.014ISI: 000371099300029ScopusID: 2-s2.0-84959521193OAI: oai:DiVA.org:kth-184949DiVA: diva2:917819
QC 201604072016-04-072016-04-072016-04-07Bibliographically approved