Stochastic neural networks (SNNs) are currently topical, with several paradigms being actively investigated including dropout, Bayesian networks, variational information bottleneck (VIB) and noise regularized learning. These network variants impact major considerations, generalization, compression, robustness against adversarial attack label noise, model calibration. However, many existing comp...

When trying to maximize the adoption of a behavior in population connected by social network, it is common strategize about where network seed behavior, often with an element randomness. Selecting seeds uniformly at random basic but compelling strategy that distributes broadly throughout network. A more sophisticated stochastic strategy, one-hop targeting, select neighbors individuals; this exp...

Stochastic activity networks (SANs) are a powerful and flexible extension of Petri nets. However, SANs facilities are not enough for modeling and analysis of complex and large-scale systems. In order to provide such high-level modeling facilities, we have defined some new extensions for SANs. We intend to preserve all capabilities of SANs for performance, dependability and performability evalua...