We show how a preselection of hidden variables can be used to efficiently train generative models with binary hidden variables. The approach is based on Expectation Maximization (...
Monitoring applications play an increasingly important role in many domains. They detect events in monitored systems and take actions such as invoke a program or notify an adminis...
materialized view selection is a non-trivial task. Hence, its complexity must be reduced. A judicious choice of views must be costdriven and influenced by the workload experienced...
Abstract. We propose a new technique called Differential Cluster Analysis for side-channel key recovery attacks. This technique uses cluster analysis to detect internal collisions ...
Abstract. This paper proposes a general approach named ExpectationMiniMax (EMM) for clustering analysis without knowing the cluster number. It describes the contrast function of Ex...