The U.S. Government Protection Profile for Separation Kernels in Environments Requiring High Robustness (SKPP) is undergoing evaluation. During its authoring process, new extended functional and assurance requirements were introduced to resolve assurance issues associated with TOE hardware, trusted initialization and trusted recovery. For high robustness, domain separation and self-protection a...

Initialization of parameters in deep neural networks has been shown to have a big impact on the performance of the networks (Mishkin & Matas, 2015). The initialization scheme devised by He et al, allowed convolution activations to carry a constrained mean which allowed deep networks to be trained effectively (He et al., 2015a). Orthogonal initializations and more generally orthogonal matrices i...

Rigorous techniques for consistent initialization of differential–algebraic systems (DAEs) are in general difficult to apply. Even worse, since initialization algorithms generally require the solution of a set of non-linear algebraic equations, they may not even converge. Hence, simpler approaches for initialization are of considerable interest in the numerical implementation of DAE models. In ...

Mutual dependencies between objects arise frequently in programs, and programmers must typically resort to manually filling “initialization holes” to help construct the corresponding object graphs, i.e. null values and/or explicitly mutable locations. This report describes a “base-line” proposal for a generalized form of value recursion in an ML-like language called initialization graphs, where...