保护定律是理解，表征和建模非线性动力系统的关键理论和实用工具。但是，对于许多复杂的动态系统，难以识别相应的保守量，因此很难分析其动力学并建立高效，稳定的预测模型。当前发现保护定律的方法通常取决于详细的动态信息，例如运动方程或细粒度的时间测量，许多最新的建议还依赖于黑匣子参数深度学习方法。相反，我们将这项任务重新制定为一种多种学习问题，并提出了一种非参数方法，将最佳运输中的Wasserstein指标与扩散图相结合，以发现从动力学系统中采样的轨迹中变化的保守数量。我们在各种物理系统上测试了这种新方法$ \ unicode {x2014} $，包括保守的汉密尔顿系统，耗散系统和时空系统$ \ unicode {x2014} $保守数量并提取其价值。使用最佳运输理论和流形学习中的工具，我们提出的方法提供了一种直接的几何方法来识别既有坚固且可解释的保护定律，而无需明确的系统模型或准确的时间信息。

符号回归是一种机器学习技术，可以学习数据的管理公式，因此有可能改变科学发现。但是，符号回归仍然受到分析系统的复杂性和维度的限制。另一方面，深度学习改变了机器学习的能力，可以分析极其复杂和高维数据集。我们提出了一个神经网络体系结构，以将符号回归扩展到参数系统，其中某些系数可能会有所不同，但是基础管理方程的结构仍然恒定。我们演示了有关各种系数的各种分析表达式，ODE和PDE的方法，并表明它可以很好地推断出训练域之外。基于神经网络的体系结构还可以与其他深度学习体系结构集成，以便在端到端训练的同时分析高维数据。为此，我们将架构与卷积神经网络集成在一起，以分析不同弹簧系统的1D图像。

识别非线性动态系统的控制方程是理解系统物理特征的关键，并构建概括超出可用数据的动态的准确模型。我们提出了一种用于发现这些管理方程的机器学习框架，仅使用部分观察，将编码器与稀疏符号模型相结合。我们的测试表明，此方法可以成功地重建完整的系统状态，并确定各种颂歌和PDE系统的底层动态。

贝叶斯优化（BO）是用于全局优化昂贵的黑盒功能的流行范式，但是在许多域中，该函数并不完全是黑色框。数据可能具有一些已知的结构（例如对称性）和/或数据生成过程可能是一个复合过程，除优化目标的值外，还可以产生有用的中间或辅助信息。但是，传统上使用的代孕模型，例如高斯工艺（GPS），随数据集大小的规模较差，并且不容易适应已知的结构。取而代之的是，我们使用贝叶斯神经网络，这是具有感应偏见的一类可扩展和灵活的替代模型，将BO扩展到具有高维度的复杂，结构化问题。我们证明了BO在物理和化学方面的许多现实问题，包括使用卷积神经网络对光子晶体材料进行拓扑优化，以及使用图神经网络对分子进行化学性质优化。在这些复杂的任务上，我们表明，就抽样效率和计算成本而言，神经网络通常优于GP作为BO的替代模型。

Deep supervised models have an unprecedented capacity to absorb large quantities of training data. Hence, training on multiple datasets becomes a method of choice towards strong generalization in usual scenes and graceful performance degradation in edge cases. Unfortunately, different datasets often have incompatible labels. For instance, the Cityscapes road class subsumes all driving surfaces, while Vistas defines separate classes for road markings, manholes etc. Furthermore, many datasets have overlapping labels. For instance, pickups are labeled as trucks in VIPER, cars in Vistas, and vans in ADE20k. We address this challenge by considering labels as unions of universal visual concepts. This allows seamless and principled learning on multi-domain dataset collections without requiring any relabeling effort. Our method achieves competitive within-dataset and cross-dataset generalization, as well as ability to learn visual concepts which are not separately labeled in any of the training datasets. Experiments reveal competitive or state-of-the-art performance on two multi-domain dataset collections and on the WildDash 2 benchmark.

In this paper, we present the findings of various methodologies for measuring the similarity of synthetic data generated from tabular data samples. We particularly apply our research to the case where the synthetic data has many more samples than the real data. This task has a special complexity: validating the reliability of this synthetically generated data with a much higher number of samples than the original. We evaluated the most commonly used global metrics found in the literature. We introduced a novel approach based on the data's topological signature analysis. Topological data analysis has several advantages in addressing this latter challenge. The study of qualitative geometric information focuses on geometric properties while neglecting quantitative distance function values. This is especially useful with high-dimensional synthetic data where the sample size has been significantly increased. It is comparable to introducing new data points into the data space within the limits set by the original data. Then, in large synthetic data spaces, points will be much more concentrated than in the original space, and their analysis will become much more sensitive to both the metrics used and noise. Instead, the concept of "closeness" between points is used for qualitative geometric information. Finally, we suggest an approach based on data Eigen vectors for evaluating the level of noise in synthetic data. This approach can also be used to assess the similarity of original and synthetic data.

Road-safety inspection is an indispensable instrument for reducing road-accident fatalities contributed to road infrastructure. Recent work formalizes road-safety assessment in terms of carefully selected risk factors that are also known as road-safety attributes. In current practice, these attributes are manually annotated in geo-referenced monocular video for each road segment. We propose to reduce dependency on tedious human labor by automating recognition with a two-stage neural architecture. The first stage predicts more than forty road-safety attributes by observing a local spatio-temporal context. Our design leverages an efficient convolutional pipeline, which benefits from pre-training on semantic segmentation of street scenes. The second stage enhances predictions through sequential integration across a larger temporal window. Our design leverages per-attribute instances of a lightweight bidirectional LSTM architecture. Both stages alleviate extreme class imbalance by incorporating a multi-task variant of recall-based dynamic loss weighting. We perform experiments on the iRAP-BH dataset, which involves fully labeled geo-referenced video along 2,300 km of public roads in Bosnia and Herzegovina. We also validate our approach by comparing it with the related work on two road-scene classification datasets from the literature: Honda Scenes and FM3m. Experimental evaluation confirms the value of our contributions on all three datasets.

Temporal data like time series are often observed at irregular intervals which is a challenging setting for existing machine learning methods. To tackle this problem, we view such data as samples from some underlying continuous function. We then define a diffusion-based generative model that adds noise from a predefined stochastic process while preserving the continuity of the resulting underlying function. A neural network is trained to reverse this process which allows us to sample new realizations from the learned distribution. We define suitable stochastic processes as noise sources and introduce novel denoising and score-matching models on processes. Further, we show how to apply this approach to the multivariate probabilistic forecasting and imputation tasks. Through our extensive experiments, we demonstrate that our method outperforms previous models on synthetic and real-world datasets.

生物学和人造药物需要处理现实世界中的不断变化。我们在四个经典的连续控制环境中研究了这个问题，并通过形态扰动增强。当不同身体部位的长度和厚度变化时，学习势头是挑战性的，因为需要控制政策才能适应形态以成功平衡和推进代理。我们表明，基于本体感受状态的控制策略的表现差，可以通过高度可变的身体配置，而（甲骨文）代理可以访问学习扰动的编码的（甲骨文）的性能要好得多。我们介绍了DMAP，这是一种以生物学启发的，基于注意力的策略网络体系结构。 DMAP将独立的本体感受处理，分布式策略与每个关节的单个控制器以及注意力机制结合在一起，从不同身体部位到不同控制器的动态门感觉信息。尽管无法访问（隐藏的）形态信息，但在所有考虑的环境中，DMAP都可以端对端训练，整体匹配或超越了Oracle代理的性能。因此，DMAP是从生物运动控制中实施原理的，为学习挑战的感觉运动任务提供了强烈的诱导偏见。总体而言，我们的工作证实了这些原则在挑战运动任务中的力量。

学习多样化的技能是机器人技术的主要挑战之一。为此，模仿学习方法取得了令人印象深刻的结果。这些方法需要明确标记的数据集或采用一致的技能执行，以使学习和积极控制单个行为，从而限制其适用性。在这项工作中，我们提出了一种合作的对抗方法，用于从未标记的数据集中获得可控技能的单一多功能策略，该数据集包含各种状态过渡模式，通过最大化其可区分性。此外，我们表明，通过在生成的对抗性模仿学习框架中利用无监督的技能发现，新颖而有用的技能随着成功的任务实现而出现。最后，在示威中编码的各种技能的忠实复制中，对获得的多功能策略进行了测试，并呈现了忠实的复制。