Bar-Ilan University

About: Bar-Ilan University is a education organization based out in Ramat Gan, Israel. It is known for research contribution in the topics: Population & Poison control. The organization has 12835 authors who have published 34964 publications receiving 995648 citations. The organization is also known as: Bar Ilan University & BIU.

Training deep neural-networks using a noise adaptation layer

Abstract: The availability of large datsets has enabled neural networks to achieve impressive recognition results. However, the presence of inaccurate class labels is known to deteriorate the performance of even the best classifiers in a broad range of classification problems. Noisy labels also tend to be more harmful than noisy attributes. When the observed label is noisy, we can view the correct label as a latent random variable and model the noise processes by a communication channel with unknown parameters. Thus we can apply the EM algorithm to find the parameters of both the network and the noise and to estimate the correct label. In this study we present a neural-network approach that optimizes the same likelihood function as optimized by the EM algorithm. The noise is explicitly modeled by an additional softmax layer that connects the correct labels to the noisy ones. This scheme is then extended to the case where the noisy labels are dependent on the features in addition to the correct labels. Experimental results demonstrate that this approach outperforms previous methods.

Current status and future directions of multivalent metal-ion batteries

Abstract: Batteries based on multivalent metals have the potential to meet the future needs of large-scale energy storage, due to the relatively high abundance of elements such as magnesium, calcium, aluminium and zinc in the Earth’s crust. However, the complexity of multivalent metal-ion chemistries has led to rampant confusions, technical challenges, and eventually doubts and uncertainties about the future of these technologies. In this Review, we clarify the key strengths as well as common misconceptions of multivalent metal-based batteries. We then examine the growth behaviour of metal anodes, which is crucial for their safety promises but hitherto unestablished. We further discuss scrutiny of anode efficiency and cathode storage mechanism pertaining to complications arising from electrolyte solutions. Finally, we critically review existing cathode materials and discuss design strategies to enable genuine multivalent metal-ion-based energy storage materials with competitive performance. Batteries based on multivalent metal anodes hold great promise for large-scale energy storage but their development is still at an early stage. This Review surveys the main complexity arising from anodes, electrolytes and cathodes, and offers views on the progression path of these technologies.

Magnetic relaxation in high-temperature superconductors

Abstract: We review experimental studies of the time decay of the nonequilibrium magnetization in high-temperature superconductors, a phenomenon known as magnetic relaxation. This effect has its origin in motion of flux lines out of their pinning sites due to thermal activation or quantum tunneling. The combination of relatively weak flux pinning and high temperatures leads to rich properties that are unconventional in the context of low temperature superconductivity and that have been the subject to intense studies. The results are assessed from a purely experimental perspective and discussed in the context of present phenomenological theories. [S0034-6861(96)00403-5]

Critical Behavior at the Onset of k → -Space Instability on the λ Line

Abstract: We calculate the critical behavior of systems having a multicritical point of a new type, hereafter called a Lifshitz point, which separates ordered phases with $\stackrel{\ensuremath{\rightarrow}}{\mathrm{k}}=0$ and $\stackrel{\ensuremath{\rightarrow}}{\mathrm{k}}\ensuremath{ e}0$ along the $\ensuremath{\lambda}$ line. For anisotropic systems, the correlation function is described in terms of four critical exponents, whereas for isotropic systems two exponents suffice. Critical exponents are calculated using an $\ensuremath{\epsilon}$-type expansion.

In Vivo Anomalous Diffusion and Weak Ergodicity Breaking of Lipid Granules

Abstract: Combining extensive single particle tracking microscopy data of endogenous lipid granules in living fission yeast cells with analytical results we show evidence for anomalous diffusion and weak ergodicity breaking. Namely we demonstrate that at short times the granules perform subdiffusion according to the laws of continuous time random walk theory. The associated violation of ergodicity leads to a characteristic turnover between two scaling regimes of the time averaged mean squared displacement. At longer times the granule motion is consistent with fractional Brownian motion.