Valence and L-shell photoionization of Cl-like argon using R-matrix techniques

Photoionization cross-sections are obtained using the relativistic DiracAtomic R-matrix Codes (DARC) for all valence and L-shell energy ranges between 27 and 270 eV. A total of 557 levels arising from the dominant configurations 3s²3p⁴, 3s³p5, 3p⁶, 3s²3p³[3d, 4s, 4p], 3p⁵3d, 3s²3p²3d², 3s³p4³d, 3s³p3³d2 and 2s²2p⁵3s²3p⁵ have been included in the targetwavefunction representation of the Ar III ion, including up to 4p in the orbital basis. We also performed a smaller Breit-Pauli (BP) calculation containing the lowest 124 levels. Direct comparisons are made with previous theoretical and experimental work for both valence shell and L-shell photoionization. Excellent agreement was found for transitions involving the ²P^o initial state to all allowed final states for both calculations across a range of photon energies. A number of resonant states have been identified to help analyse and explain the nature of the spectra at photon energies between 250 and 270 eV.

Electron-impact excitation of open d-shell ions

Astrophysics is driven by observations, and in the present era there are a wealth of state-of-the-art ground-based and satellite facilities. The astrophysical spectra emerging from these are of exceptional quality and quantity and cover a broad wavelength range. To meaningfully interpret these spectra, astronomers employ highly complex modelling codes to simulate the astrophysical observations. Important input to these codes include atomic data such as excitation rates, photoionization cross sections, oscillator strengths, transition probabilities and energy levels/line wavelengths. Due to the relatively low temperatures associated with many astrophysical plasmas, the accurate determination of electron-impact excitation rates in the low energy region is essential in generating a reliable spectral synthesis. Hence it is these atomic data, and the main computational methods used to evaluate them, which we focus on in this publication. We consider in particular the complicated open d- shell structures of the Fe-peak ions in low ionization stages. While some of these data can be obtained experimentally, they are usually of insufficient accuracy or limited to a small number of transitions.

Comparison of Parallelised R-matrix codes for electron impact excitation of Cr II

A comparison of collision strengths and effective collision strengths has been undertaken for the Cr II ion based on the model of Wasson et al [2010 A & A. 524 A35]. Calculations have been completed using the Breit-Pauli, RMATRX II and DARC suites of codes.

Analgesic exposure in pregnant rats affects fetal germ cell development with inter-generational reproductive consequences

Analgesics which affect prostaglandin (PG) pathways are used by most pregnant women. As germ cells (GC) undergo developmental and epigenetic changes in fetal life and are PG targets, we investigated if exposure of pregnant rats to analgesics (indomethacin or acetaminophen) affected GC development and reproductive function in resulting offspring (F1) or in the F2 generation. Exposure to either analgesic reduced F1 fetal GC number in both sexes and altered the tempo of fetal GC development sex-dependently, with delayed meiotic entry in oogonia but accelerated GC differentiation in males. These effects persisted in adult F1 females as reduced ovarian and litter size, whereas F1 males recovered normal GC numbers and fertility by adulthood. F2 offspring deriving from an analgesic-exposed F1 parent also exhibited sex-specific changes. F2 males exhibited normal reproductive development whereas F2 females had smaller ovaries and reduced follicle numbers during puberty/adulthood; as similar changes were found for F2 offspring of analgesic-exposed F1 fathers or mothers, we interpret this as potentially indicating an analgesic-induced change to GC in F1. Assuming our results are translatable to humans, they raise concerns that analgesic use in pregnancy could potentially affect fertility of resulting daughters and grand-daughters.

Learning Bayesian Networks with Bounded Tree-Width via Guided Search

Bounding the tree-width of a Bayesian network can reduce the chance of overfitting, and allows exact inference to be performed efficiently. Several existing algorithms tackle the problem of learning bounded tree-width Bayesian networks by learning from k-trees as super-structures, but they do not scale to large domains and/or large tree-width. We propose a guided search algorithm to find k-trees with maximum Informative scores, which is a measure of quality for the k-tree in yielding good Bayesian networks. The algorithm achieves close to optimal performance compared to exact solutions in small domains, and can discover better networks than existing approximate methods can in large domains. It also provides an optimal elimination order of variables that guarantees small complexity for later runs of exact inference. Comparisons with well-known approaches in terms of learning and inference accuracy illustrate its capabilities.

Statistical Tests for Joint Analysis of Performance Measures

Recently there has been an increasing interest in the development of new methods using Pareto optimality to deal with multi-objective criteria (for example, accuracy and architectural complexity). Once one has learned a model based on their devised method, the problem is then how to compare it with the state of art. In machine learning, algorithms are typically evaluated by comparing their performance on different data sets by means of statistical tests. Unfortunately, the standard tests used for this purpose are not able to jointly consider performance measures. The aim of this paper is to resolve this issue by developing statistical procedures that are able to account for multiple competing measures at the same time. In particular, we develop two tests: a frequentist procedure based on the generalized likelihood-ratio test and a Bayesian procedure based on a multinomial-Dirichlet conjugate model. We further extend them by discovering conditional independences among measures to reduce the number of parameter of such models, as usually the number of studied cases is very reduced in such comparisons. Real data from a comparison among general purpose classifiers is used to show a practical application of our tests.

Local Computation in Credal Networks

The goal of this contribution is to discuss local computation in credal networks — graphical models that can represent imprecise and indeterminate probability values. We analyze the inference problem in credal networks, discuss how inference algorithms can benefit from local computation, and suggest that local computation can be particularly important in approximate inference algorithms.