The interaction of relativistically intense lasers with opaque targets represents a highly non-linear, multi-dimensional parameter space. This limits the utility of sequential 1D scanning of experimental parameters for the optimization of secondary radiation, although to-date this has been the accepted methodology due to low data acquisition rates. High repetition-rate (HRR) lasers augmented by machine learning present a valuable opportunity for efficient source optimization. Here, an automated, HRR-compatible system produced high-fidelity parameter scans, revealing the influence of laser intensity on target pre-heating and proton generation. A closed-loop Bayesian optimization of maximum proton energy, through control of the laser wavefront and target position, produced proton beams with equivalent maximum energy to manually optimized laser pulses but using only 60% of the laser energy. This demonstration of automated optimization of laser-driven proton beams is a crucial step towards deeper physical insight and the construction of future radiation sources.

We present the development and characterization of a high-stability, multi-material, multi-thickness tape-drive target for laser-driven acceleration at repetition rates of up to 100 Hz. The tape surface position was measured to be stable on the sub-micrometre scale, compatible with the high-numerical aperture focusing geometries required to achieve relativistic intensity interactions with the pulse energy available in current multi-Hz and near-future higher repetition-rate lasers ($>$kHz). Long-term drift was characterized at 100 Hz demonstrating suitability for operation over extended periods. The target was continuously operated at up to 5 Hz in a recent experiment for 70,000 shots without intervention by the experimental team, with the exception of tape replacement, producing the largest data-set of relativistically intense laser–solid foil measurements to date. This tape drive provides robust targetry for the generation and study of high-repetition-rate ion beams using next-generation high-power laser systems, also enabling wider applications of laser-driven proton sources.

A machine learning model was created to predict the electron spectrum generated by a GeV-class laser wakefield accelerator. The model was constructed from variational convolutional neural networks, which mapped the results of secondary laser and plasma diagnostics to the generated electron spectrum. An ensemble of trained networks was used to predict the electron spectrum and to provide an estimation of the uncertainty of that prediction. It is anticipated that this approach will be useful for inferring the electron spectrum prior to undergoing any process that can alter or destroy the beam. In addition, the model provides insight into the scaling of electron beam properties due to stochastic fluctuations in the laser energy and plasma electron density.

Previous studies have shown that schizophrenic patients are more likely to be born in winter or early spring months than the general population. Data on 4,207 patients with a hospital diagnosis of schizophrenia were obtained from a mailed survey to public departments of adult psychiatry in metropolitan France. For each year from 1900 to 1965, the expected monthly number of schizophrenic births was calculated and any seasonal variation of live births in the general population was taken into account. Cumulative distributions of the observed and expected number of schizophrenic births were compared using a Kolmogorov-Smirnov type statistic. The seasonal distribution of schizophrenic births was significantly different from that of the general population (P < 0.01). An excess of schizophrenic births was found in the first half of the year, with a peak in April (+ 13%).

The purpose of this study was to determine cross-resistance patterns among wild oat lines resistant to acetyl-CoA carboxylase (ACCase) inhibitors and to determine which, if any, cross-resistant type was more common than another. Discriminatory concentrations of two aryloxyphenoxy-propionates (APP) and three cyclohexanediones (CHD) were determined using a petri-dish bioassay. These concentrations were then applied to 82 resistant wild oat lines identified in previous studies. In addition, two resistant standards (UM1 and UM33) and a susceptible standard (UM5) were included in the experiments. Coleoptile lengths expressed as percentages of untreated controls were used to assess the level of resistance to each herbicide. Large variations were observed among wild oat lines and herbicides. However, cluster analysis summarized the relationship between the five herbicides (variables) and the wild oat lines into three main cross-resistance types. Type A included wild oat lines with high resistance to APP herbicides and no or low resistance to CHD herbicides. Types B and C included those with low to moderate resistant and high levels of resistance to all five herbicides, respectively. Type C was the most common cross-resistance type. Relationships among herbicides were determined using pairwise correlation and principal component analysis (PCA). All correlations were high between APP herbicides and between CHD herbicides but not between APP and CHD herbicides. The first two axes of the PCA accounted for 88.4% of the total variance, with the first axis correlated to the CHD herbicides and the second axis correlated to the APP herbicides. In the PCA, wild oat lines were segregated into the three types identified in the cluster analysis. Although CHD and APP herbicides bind at the same region on the ACCase, resistant wild oat lines respond differently to them.