LATIDIA · Ciberseguridad

El tiempo lo es todo: inferir membresía temporal en modelos continuamente actualizados

arXiv: 2611.05781v1Tipo de anuncio: nuevo Resumen: La actualización continua del modelo se ha convertido en una práctica común en el despliegue de sistemas de aprendizaje automático. Sin embargo, las conductas dinámicas de entrenamiento dan lugar a nuevos riesgos que no sufren

La noticia

arXiv:2611.05781v1 Announce Type: new Abstract: Continual model updating has become a common practice in the deployment of machine learning systems. However, the dynamic training behaviors give rise to new risks not sufficiently addressed by existing studies. In this paper, we present the first systematic study of temporal privacy risk in continually updated deep learning models. We formalize the notion of temporal membership in continually updated models. We then define the temporal membership inference attack, which seeks to infer whether and when individual samples were involved in model training, at both timestamp-level and interval-level granularities. To this end, we propose TMIA, a novel black-box attack that leverages bidirectional LSTMs to infer