Previous work has shown the potential of deep learning to predict renal obstruction using kidney ultrasound images. However, these image-based classifiers have been trained with the goal of single-visit inference in mind. We compare methods from video action recognition (i.e. convolutional pooling, LSTM, TSM) to adapt single-visit convolutional models to handle multiple visit inference. We demonstrate that incorporating images from a patient's past hospital visits provides only a small benefit for the prediction of obstructive hydronephrosis. Therefore, inclusion of prior ultrasounds is beneficial, but prediction based on the latest ultrasound is sufficient for patient risk stratification.

Automated synthesis of provably correct controllers for cyber-physical systems is crucial for deploying these systems in safety-critical scenarios. However, their hybrid features and stochastic or unknown behaviours make this synthesis problem challenging. In this paper, we propose a method for synthesizing controllers for Markov jump linear systems (MJLSs), a particular class of cyber-physical systems, that certifiably satisfy a requirement expressed as a specification in probabilistic computation tree logic (PCTL). An MJLS consists of a finite set of linear dynamics with unknown additive disturbances, where jumps between these modes are governed by a Markov decision process (MDP). We consider both the case where the transition function of this MDP is given by probability intervals or where it is completely unknown. Our approach is based on generating a finite-state abstraction which captures both the discrete and the continuous behaviour of the original system. We formalise such abstraction as an interval Markov decision process (iMDP): intervals of transition probabilities are computed using sampling techniques from the so-called "scenario approach", resulting in a probabilistically sound approximation of the MJLS. This iMDP abstracts both the jump dynamics between modes, as well as the continuous dynamics within the modes. To demonstrate the efficacy of our technique, we apply our method to multiple realistic benchmark problems, in particular, temperature control, and aerial vehicle delivery problems.

我们介绍了DeepGen，这是一个在网络范围内部署的系统，用于自动为宾果派客户创建赞助的搜索广告（ADS）。我们利用最新的自然语言生成（NLG）模型以抽象的方式从广告商的网页中生成流利的广告，并解决了实际问题，例如事实和推理速度。此外，我们的系统可实时创建自定义的广告，以响应用户的搜索查询，因此根据用户所需的内容突出显示了同一产品的不同方面。为了实现这一目标，我们的系统会提前生成各种较小广告的选择，并在查询时间选择最相关的广告选择，以将其缝合为完整的广告。我们通过培训可控的NLG模型来改善发电多样性，以生成相同网页的多个广告，突出显示不同的销售点。我们的系统设计通过首先运行具有不同目标训练的生成模型的合奏，然后使用多样性采样算法来选择各种各样的生成结果以进行在线选择，从而进一步改善了多样性。实验结果显示了我们提出的系统设计的有效性。我们的系统目前已在生产中部署，为Bing提供的全球广告提供$ {\ sim} 4 \％$。

联合学习（FL）是数据是私人且敏感时的有前途的分布式学习框架。但是，当数据是异质且非独立且相同分布的（非IID）时，此框架中最新的解决方案并不是最佳的。我们提出了一种实用且强大的佛罗里达州个性化方法，该方法通过平衡探索和利用几种全球模型来适应异质和非IID数据。为了实现我们的个性化目标，我们使用了专家（MOE）的混合，这些专家（MOE）学会了分组彼此相似的客户，同时更有效地使用全球模型。我们表明，与病理非IID环境中的本地模型相比，我们的方法的准确性高达29.78％，高达4.38％，即使我们在IID环境中调整了方法。

A common assumption in causal inference from observational data is that there is no hidden confounding. Yet it is, in general, impossible to verify the presence of hidden confounding factors from a single dataset. Under the assumption of independent causal mechanisms underlying the data generating process, we demonstrate a way to detect unobserved confounders when having multiple observational datasets coming from different environments. We present a theory for testable conditional independencies that are only absent during hidden confounding and examine cases where we violate its assumptions: degenerate & dependent mechanisms, and faithfulness violations. Additionally, we propose a procedure to test these independencies and study its empirical finite-sample behavior using simulation studies and semi-synthetic data based on a real-world dataset. In most cases, our theory correctly predicts the presence of hidden confounding, particularly when the confounding bias is~large.

最近的工作表明，（1）增加输入长度或（2）增加模型大小可以提高基于变压器的神经模型的性能。在本文中，我们提出了一个名为Longt5的新模型，我们探讨了同时缩放输入长度和模型大小的效果。具体而言，我们综合了从长输入变压器（ETC）的关注思路，并采用了从摘要预训练（PEGASU）的预训练策略进入可扩展的T5架构。结果是我们称之为{\ EM瞬态全球}（TGLOBAL）的新关注机制，这些机制是模仿等本地/全球注意力机制，但不需要额外的侧面输入。我们能够实现最先进的结果，以若干摘要任务，优于问题应答任务的原始T5模型。

高分辨率气象雷达图像的可用性是有效的预测和决策。在超越传统雷达覆盖范围之外，生成模型已成为一种重要的合成能力，融合更普遍的数据来源，例如卫星图像和数值天气模型，进入准确的雷达样产品。在这里，我们展示了使用量子辅助模型来增强传统卷积神经网络的方法，用于全球合成天气雷达中的生成任务。我们表明Quantum Kernels原则上可以根据相关底层数据上的古典学习机来表现出基本上更复杂的任务。我们的结果建立了合成气象雷达作为量子计算能力的有效启发式基准，并在高影响力的相关问题上设定了详细量子优势基准测试的阶段。

Surrogate algorithms such as Bayesian optimisation are especially designed for black-box optimisation problems with expensive objectives, such as hyperparameter tuning or simulation-based optimisation. In the literature, these algorithms are usually evaluated with synthetic benchmarks which are well established but have no expensive objective, and only on one or two real-life applications which vary wildly between papers. There is a clear lack of standardisation when it comes to benchmarking surrogate algorithms on real-life, expensive, black-box objective functions. This makes it very difficult to draw conclusions on the effect of algorithmic contributions and to give substantial advice on which method to use when. A new benchmark library, EXPObench, provides first steps towards such a standardisation. The library is used to provide an extensive comparison of six different surrogate algorithms on four expensive optimisation problems from different real-life applications. This has led to new insights regarding the relative importance of exploration, the evaluation time of the objective, and the used model. We also provide rules of thumb for which surrogate algorithm to use in which situation. A further contribution is that we make the algorithms and benchmark problem instances publicly available, contributing to more uniform analysis of surrogate algorithms. Most importantly, we include the performance of the six algorithms on all evaluated problem instances. This results in a unique new dataset that lowers the bar for researching new methods as the number of expensive evaluations required for comparison is significantly reduced.

虽然非自杀自我受伤（NSSI）不是一个新的现象，但仍有有限的有限仍然符合对行为的理解，行为背后的意图以及个人自己对他们的行为所说的。本研究从Pro-NSSI上的Reddit收集了Pro-NSSI公共博客帖子，并使用LIWC软件对内容进行了分析，以检查NSSI特定词，语言特性和心理语言学属性的使用。检查了。结果通过消除神话来告知现有的咨询实践，并向参与使用NSSII的人们的内心世界提供洞察力。在Reddit博客中，对于NSSI特定单词类别的最常见类别的类别，是Reddit博客中最常使用的NSSI的一个接触的原因。分析中发现的语言特性反映了预测结果; Pro-NSSI帖子的作者使用广泛展示了一人奇异代词的预期结果，这表明高水平的心理健康窘迫和隔离。在这些公共Reddit帖子中可以观察到的心理语言学属性在负面情绪中，展示了青年和冲动的负面情绪。当分析这些帖子时发现的语言特性支持早期研究的工作，这些研究剥夺了在心理健康界中流传的NSSI的常见神话。这些调查结果表明，参与NSSI的人的语言支持在宣布关于NSSI的常见神话中的研究发现。