Hoeffding (1948a) developed the basic theory of U-Statistics, a family of estimates which includes many familiar and interesting examples. This lecture reviews this theory. Standard references for the material presented here include Serfling (1980, Chapter 5), Lehmman (1999, Chapter 6) and van der Vaart (1998, Chapters 11 & 12). The basic theory of U-Statistics allows for a presentation of larg...

A technique is described for the simultaneous and controlled random mutation of all three heavy or light chain complementaritj-determining regions (CDRs) in a single-chain Fv specific for the 0 polysaccharide of SalmoneUla serogroup B. Sense oligonucleotides were synthesized such that the central bases encoding a CDR were randomized by equimolar spiking with A, G, C, and T at a level of 10%/1 w...

The isotopic composition of U in nature is generally assumed to be invariant. Here, we report variations of the U/U isotope ratio in natural samples (basalts, granites, seawater, corals, black shales, suboxic sediments, ferromanganese crusts/ nodules and BIFs) of 1.3‰, exceeding by far the analytical precision of our method ( 0.06‰, 2SD). U isotopes were analyzed with MC-ICP-MS using a mixed U–...

As deep learning algorithms are widely adopted, an increasing number of them positioned in embedded application domains with strict reliability constraints. The expenditure significant resources to satisfy performance requirements neural network accelerators has thinned out the margins for delivering safety applications, thus precluding adoption conventional fault tolerance methods. potential e...

We present a dynamic network rewiring (DNR) method to generate pruned deep neural (DNN) models that both are robust against adversarially generated images and maintain high accuracy on clean images. In particular, the disclosed DNR training is based unified constrained optimization formulation using novel hybrid loss function merges sparse learning with adversarial training. This strategy dynam...