Search

We consider entanglement-assisted communication over the qubit depolarizing channel under the security requirement of covert communication, where not only the information is kept secret, but the transmission itself must be concealed from detection by an adversary. Previous work showed that $O(\sqrt{n})$ information bits can be reliably and covertly transmitted in $n$ channel uses without entanglement assistance. However, Gagatsos et al. (2020) showed that entanglement assistance can increase this scaling to $O(\sqrt{n}\log(n))$ for continuous-variable bosonic channels. Here, we present a finite-dimensional parallel, and show that $O(\sqrt{n}\log(n))$ covert bits can be transmitted reliably over $n$ uses of a qubit depolarizing channel.

Classifiers and other statistics-based machine learning (ML) techniques generalize, or learn, based on various statistical properties of the training data. The assumption underlying statistical ML resulting in theoretical or empirical performance guarantees is that the distribution of the training data is representative of the production data distribution. This assumption often breaks; for instance, statistical distributions of the data may change. We term changes that affect ML performance `data drift' or `drift'. Many classification techniques compute a measure of confidence in their results. This measure might not reflect the actual ML performance. A famous example is the Panda picture that is correctly classified as such with a confidence of about 60\%, but when noise is added it is incorrectly classified as a Gibbon with a confidence of above 99\%. However, the work we report on here suggests that a classifier's measure of confidence can be used for the purpose of detecting data drift. We propose an approach based solely on classifier suggested labels and its confidence in them, for alerting on data distribution or feature space changes that are likely to cause data drift. Our approach identities degradation in model performance and does not require labeling of data in production which is often lacking or delayed. Our experiments with three different data sets and classifiers demonstrate the effectiveness of this approach in detecting data drift. This is especially encouraging as the classification itself may or may not be correct and no model input data is required. We further explore the statistical approach of sequential change-point tests to automatically determine the amount of data needed in order to identify drift while controlling the false positive rate (Type-1 error).

The objective of the Geopotential Panel was to develop a program of data acquisition and model development for the Earth's gravity and magnetic fields that meet the basic science requirements of the solid Earth and ocean studies. Presented here are the requirements for gravity information and models through the end of the century, the present status of our knowledge, data acquisition techniques, and an outline of a program to meet the requirements.

Spanwise lift distributions have been calculated for nineteen unswept wings with various aspect ratios and taper ratios and with a variety of angle-of-attack or twist distributions, including flap and aileron deflections, by means of the Weissinger method with eight control points on the semispan. Also calculated were aerodynamic influence coefficients which pertain to a certain definite set of stations along the span, and several methods are presented for calculating aerodynamic influence functions and coefficients for stations other than those stipulated. The information presented in this report can be used in the analysis of untwisted wings or wings with known twist distributions, as well as in aeroelastic calculations involving initially unknown twist distributions.