Abstract: Structure learning of dynamic Bayesian networks provide a principled mechanism for identifying conditional dependencies in time-series data. This learning procedure assum...
: To understand overall vulnerability to network attack, one must consider attacker exploits not just in isolation, but also in combination. That is, one must analyze how low-level...
A practical lipreading system can be considered either as subject dependent (SD) or subject-independent (SI). An SD system is user-specific, i.e., customized for some particular ...
Abstract. Graph games of infinite length provide a natural model for open reactive systems: one player (Eve) represents the controller and the other player (Adam) represents the en...
This work analyzes the connectivity of large diameter networks where every link has an independent probability p of failure. We give a (relatively simple) topological condition th...