Simulated annealing in the presence of noise

In many practical optimization problems, evaluation of a solution is subject to noise, e.g., due to stochastic simulations or measuring errors. Therefore, heuristics are needed that are capable of handling such noise. This paper first reviews the state-of-the-art in applying simulated annealing to noisy optimization problems. Then, two new algorithmic variants are proposed: an improved version of stochastic annealing that allows for arbitrary annealing schedules, and a new approach called simulated annealing in noisy environments (SANE). The latter integrates ideas from statistical sequential selection in order to reduce the number of samples required for making an acceptance decision with sufficient statistical confidence. Finally, SANE is shown to significantly outperform other state-of-the-art simulated annealing techniques on a stochastic travelling salesperson problem.     

Is Nd2−xCexCuO4 an electron-superconductor?

We present results from electron energy-loss spectroscopy on Nd_2–xCe_xCuO₄. There is no clear evidence for Cu¹⁺ but for holes at the oxygen sites. The valency of Ce is found to be close to 3.4. Thus our results do not corroboraten-type superconductivity in this compound.     

Ground state properties of a spinless Falicov-Kimball model; additional features

We have studied some energetic and structural ground state properties of a spinless Falicov-Kimball model. Using a method based on “Tchebycheff-Markoff inequalities”, we have calculated sharp upper and lower bounds for the ground state energy. These calculations lead to rigorous results for a special range of parameters, where we are able to give the exactf-level occupation pattern for a square lattice. The results for the “symmetric” case, where a superstructure occurs leading to a metal-insulator transition, have already been published. In this paper we present some additional results also for the “unsymmetric case” and study valence-transitions for fixed particle-number.     

Probabilistic Predictions with Federated Learning

Probabilistic predictions with machine learning are important in many applications. These are commonly done with Bayesian learning algorithms. However, Bayesian learning methods are computationally expensive in comparison with non-Bayesian methods. Furthermore, the data used to train these algorithms are often distributed over a large group of end devices. Federated learning can be applied in this setting in a communication-efficient and privacy-preserving manner but does not include predictive uncertainty. To represent predictive uncertainty in federated learning, our suggestion is to introduce uncertainty in the aggregation step of the algorithm by treating the set of local weights as a posterior distribution for the weights of the global model. We compare our approach to state-of-the-art Bayesian and non-Bayesian probabilistic learning algorithms. By applying proper scoring rules to evaluate the predictive distributions, we show that our approach can achieve similar performance as the benchmark would achieve in a non-distributed setting.     

Lumineszenzdatierung als Schlüssel zur Vergangenheit

In this article, we describe the basic principles of luminescence dating, for which knowledge and methods of solid state physics and radiochemistry play a major role. The type of dated event – the last heating or daylight exposure – opens up a broad spectrum of applications of this method for periods ranging from 10² up to several 10⁵ years. Through the cooperation of dating specialists with geoscientists and archaeologists, luminescence dating is nowadays an essential tool for addressing problems in Earth‐science and human history.     

ChemInform Abstract: A Matrix Isolation and Computational Study of Molecular Palladium Fluorides: Does PdF6 Exist?

Binary palladium fluorides from PdF to PdF₆ are investigated by matrix‐isolation methods using thermal evaporation and laser ablation to generate Pd atoms for reaction with F₂‐doped Ar and Ne matrices as well as neat F₂ matrices.