Coralie Busse-Grawitz, ETH Zurich
There is an abundance of current security research that heavily relies on Machine Learning to find a model for a complex, not well-understood truth. However, offloading the task of finding the truth to a tool eventually relies on circular reasoning, rendering these security applications useless and dangerous: Measuring a model's quality requires a dataset, whose representativeness can only be guaranteed with thorough a priori knowledge of the truth, causing an implicit tautology. This talk pushes for a clear separation between modeling and finding the truth, to give future security systems a steady ground to stand on.
Open Access Media
USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.
title = {Talk 5},
year = {2019},
address = {Santa Clara, CA},
publisher = {USENIX Association},
month = aug
}