从数据中学到的分类器越来越多地用作安全是关键问题的系统中的组件。在这项工作中,我们通过称为安全订购约束的约束来提出针对分类器的正式安全概念。这些限制条件将分类器输出的类输出的顺序与输入的条件有关,并且表达足以编码文献中分类器安全规范的各种有趣的示例。对于使用神经网络实施的分类器,我们还提出了一种运行时机制,用于执行安全订购约束。我们的方法基于一个自我校正层,该层可证明,无论分类器输入的特征如何,它都可以产生安全的输出。我们将此层与现有的神经网络分类器组成,以构建自我校正网络(SC-NET),并证明除了提供安全的输出外,SC-NET还可以保证尽可能保留原始网络的分类精度。我们的方法独立于用于分类的神经网络的大小和体系结构,仅取决于指定的属性和网络输出的尺寸;因此,它可扩展到大型最新网络。我们表明,我们的方法可以针对GPU进行优化,从而在当前硬件上引入了少于1ms的运行时开销 - 即使在包含数十万个神经元和数百万参数的大型,广泛使用的网络上。
translated by 谷歌翻译
组合优化是运营研究和计算机科学领域的一个公认领域。直到最近,它的方法一直集中在孤立地解决问题实例,而忽略了它们通常源于实践中的相关数据分布。但是,近年来,人们对使用机器学习,尤其是图形神经网络(GNN)的兴趣激增,作为组合任务的关键构件,直接作为求解器或通过增强确切的求解器。GNN的电感偏差有效地编码了组合和关系输入,因为它们对排列和对输入稀疏性的意识的不变性。本文介绍了对这个新兴领域的最新主要进步的概念回顾,旨在优化和机器学习研究人员。
translated by 谷歌翻译
背景信息:在过去几年中,机器学习(ML)一直是许多创新的核心。然而,包括在所谓的“安全关键”系统中,例如汽车或航空的系统已经被证明是非常具有挑战性的,因为ML的范式转变为ML带来完全改变传统认证方法。目的:本文旨在阐明与ML为基础的安全关键系统认证有关的挑战,以及文献中提出的解决方案,以解决它们,回答问题的问题如何证明基于机器学习的安全关键系统?'方法:我们开展2015年至2020年至2020年之间发布的研究论文的系统文献综述(SLR),涵盖了与ML系统认证有关的主题。总共确定了217篇论文涵盖了主题,被认为是ML认证的主要支柱:鲁棒性,不确定性,解释性,验证,安全强化学习和直接认证。我们分析了每个子场的主要趋势和问题,并提取了提取的论文的总结。结果:单反结果突出了社区对该主题的热情,以及在数据集和模型类型方面缺乏多样性。它还强调需要进一步发展学术界和行业之间的联系,以加深域名研究。最后,它还说明了必须在上面提到的主要支柱之间建立连接的必要性,这些主要柱主要主要研究。结论:我们强调了目前部署的努力,以实现ML基于ML的软件系统,并讨论了一些未来的研究方向。
translated by 谷歌翻译
在过去的十年中,神经网络(NNS)已被广泛用于许多应用程序,包括安全系统,例如自主系统。尽管采用了新兴的采用,但众所周知,NNS容易受到对抗攻击的影响。因此,提供确保此类系统正常工作的保证非常重要。为了解决这些问题,我们介绍了一个修复不安全NNS W.R.T.的框架。安全规范,即利用可满足的模型理论(SMT)求解器。我们的方法能够通过仅修改其重量值的一些重量值来搜索新的,安全的NN表示形式。此外,我们的技术试图最大程度地提高与原始网络在其决策边界方面的相似性。我们进行了广泛的实验,以证明我们提出的框架能够产生安全NNS W.R.T.的能力。对抗性的鲁棒性特性,只有轻度的准确性损失(就相似性而言)。此外,我们将我们的方法与天真的基线进行比较,以证明其有效性。总而言之,我们提供了一种算法以自动修复具有安全性的算法,并建议一些启发式方法以提高其计算性能。当前,通过遵循这种方法,我们能够产生由分段线性relu激活函数组成的小型(即具有多达数百个参数)的小型(即具有多达数百个参数)。然而,我们的框架是可以合成NNS W.R.T.的一般框架。一阶逻辑规范的任何可决定片段。
translated by 谷歌翻译
Deep neural networks have emerged as a widely used and effective means for tackling complex, real-world problems. However, a major obstacle in applying them to safety-critical systems is the great difficulty in providing formal guarantees about their behavior. We present a novel, scalable, and efficient technique for verifying properties of deep neural networks (or providing counter-examples). The technique is based on the simplex method, extended to handle the non-convex Rectified Linear Unit (ReLU ) activation function, which is a crucial ingredient in many modern neural networks. The verification procedure tackles neural networks as a whole, without making any simplifying assumptions. We evaluated our technique on a prototype deep neural network implementation of the next-generation airborne collision avoidance system for unmanned aircraft (ACAS Xu). Results show that our technique can successfully prove properties of networks that are an order of magnitude larger than the largest networks verified using existing methods.
translated by 谷歌翻译
In the past few years, neural architecture search (NAS) has become an increasingly important tool within the deep learning community. Despite the many recent successes of NAS, however, most existing approaches operate within highly structured design spaces, and hence explore only a small fraction of the full search space of neural architectures while also requiring significant manual effort from domain experts. In this work, we develop techniques that enable efficient NAS in a significantly larger design space. To accomplish this, we propose to perform NAS in an abstract search space of program properties. Our key insights are as follows: (1) the abstract search space is significantly smaller than the original search space, and (2) architectures with similar program properties also have similar performance; thus, we can search more efficiently in the abstract search space. To enable this approach, we also propose a novel efficient synthesis procedure, which accepts a set of promising program properties, and returns a satisfying neural architecture. We implement our approach, $\alpha$NAS, within an evolutionary framework, where the mutations are guided by the program properties. Starting with a ResNet-34 model, $\alpha$NAS produces a model with slightly improved accuracy on CIFAR-10 but 96% fewer parameters. On ImageNet, $\alpha$NAS is able to improve over Vision Transformer (30% fewer FLOPS and parameters), ResNet-50 (23% fewer FLOPS, 14% fewer parameters), and EfficientNet (7% fewer FLOPS and parameters) without any degradation in accuracy.
translated by 谷歌翻译
即使机器学习算法已经在数据科学中发挥了重要作用,但许多当前方法对输入数据提出了不现实的假设。由于不兼容的数据格式,或数据集中的异质,分层或完全缺少的数据片段,因此很难应用此类方法。作为解决方案,我们提出了一个用于样本表示,模型定义和培训的多功能,统一的框架,称为“ Hmill”。我们深入审查框架构建和扩展的机器学习的多个范围范式。从理论上讲,为HMILL的关键组件的设计合理,我们将通用近似定理的扩展显示到框架中实现的模型所实现的所有功能的集合。本文还包含有关我们实施中技术和绩效改进的详细讨论,该讨论将在MIT许可下发布供下载。该框架的主要资产是其灵活性,它可以通过相同的工具对不同的现实世界数据源进行建模。除了单独观察到每个对象的一组属性的标准设置外,我们解释了如何在框架中实现表示整个对象系统的图表中的消息推断。为了支持我们的主张,我们使用框架解决了网络安全域的三个不同问题。第一种用例涉及来自原始网络观察结果的IoT设备识别。在第二个问题中,我们研究了如何使用以有向图表示的操作系统的快照可以对恶意二进制文件进行分类。最后提供的示例是通过网络中实体之间建模域黑名单扩展的任务。在所有三个问题中,基于建议的框架的解决方案可实现与专业方法相当的性能。
translated by 谷歌翻译
Two approaches to AI, neural networks and symbolic systems, have been proven very successful for an array of AI problems. However, neither has been able to achieve the general reasoning ability required for human-like intelligence. It has been argued that this is due to inherent weaknesses in each approach. Luckily, these weaknesses appear to be complementary, with symbolic systems being adept at the kinds of things neural networks have trouble with and vice-versa. The field of neural-symbolic AI attempts to exploit this asymmetry by combining neural networks and symbolic AI into integrated systems. Often this has been done by encoding symbolic knowledge into neural networks. Unfortunately, although many different methods for this have been proposed, there is no common definition of an encoding to compare them. We seek to rectify this problem by introducing a semantic framework for neural-symbolic AI, which is then shown to be general enough to account for a large family of neural-symbolic systems. We provide a number of examples and proofs of the application of the framework to the neural encoding of various forms of knowledge representation and neural network. These, at first sight disparate approaches, are all shown to fall within the framework's formal definition of what we call semantic encoding for neural-symbolic AI.
translated by 谷歌翻译
当前独立于域的经典计划者需要问题域和实例作为输入的符号模型,从而导致知识采集瓶颈。同时,尽管深度学习在许多领域都取得了重大成功,但知识是在与符号系统(例如计划者)不兼容的亚符号表示中编码的。我们提出了Latplan,这是一种无监督的建筑,结合了深度学习和经典计划。只有一组未标记的图像对,显示了环境中允许的过渡子集(训练输入),Latplan学习了环境的完整命题PDDL动作模型。稍后,当给出代表初始状态和目标状态(计划输入)的一对图像时,Latplan在符号潜在空间中找到了目标状态的计划,并返回可视化的计划执行。我们使用6个计划域的基于图像的版本来评估LATPLAN:8个插头,15个式嘴,Blockworld,Sokoban和两个LightsOut的变体。
translated by 谷歌翻译
最近已经提出了几个查询和分数来解释对ML模型的个人预测。鉴于ML型号的灵活,可靠和易于应用的可解释性方法,我们预见了需要开发声明语言以自然地指定不同的解释性查询。我们以原则的方式通过源于逻辑,称为箔,允许表达许多简单但重要的解释性查询,并且可以作为更具表现力解释性语言的核心来实现这一语言。我们研究箔片查询的两类ML模型的计算复杂性经常被视为容易解释:决策树和OBDD。由于ML模型的可能输入的数量是尺寸的指数,因此箔评估问题的易易性是精细的,但是可以通过限制模型的结构或正在评估的箔片段来实现。我们还以高级声明语言包装的箔片的原型实施,并执行实验,表明可以在实践中使用这种语言。
translated by 谷歌翻译
最近,图形神经网络(GNN)已应用于群集上的调整工作,比手工制作的启发式方法更好地表现了。尽管表现令人印象深刻,但仍然担心这些基于GNN的工作调度程序是否满足用户对其他重要属性的期望,例如防止策略,共享激励和稳定性。在这项工作中,我们考虑对基于GNN的工作调度程序的正式验证。我们解决了几个特定领域的挑战,例如网络,这些挑战比验证图像和NLP分类器时遇到的更深层和规格更丰富。我们开发了拉斯维加斯,这是基于精心设计的算法,将这些调度程序的单步和多步属性验证的第一个通用框架,它们结合了抽象,改进,求解器和证明传输。我们的实验结果表明,与以前的方法相比,维加斯在验证基于GNN的调度程序的重要特性时会达到显着加速。
translated by 谷歌翻译
归纳逻辑编程(ILP)是一种机器学习的形式。ILP的目标是诱导推广培训示例的假设(一组逻辑规则)。随着ILP转30,我们提供了对该领域的新介绍。我们介绍了必要的逻辑符号和主要学习环境;描述ILP系统的构建块;比较几个维度的几个系统;描述四个系统(Aleph,Tilde,Aspal和Metagol);突出关键应用领域;最后,总结了未来研究的当前限制和方向。
translated by 谷歌翻译
Deep neural networks have achieved impressive experimental results in image classification, but can surprisingly be unstable with respect to adversarial perturbations, that is, minimal changes to the input image that cause the network to misclassify it. With potential applications including perception modules and end-to-end controllers for self-driving cars, this raises concerns about their safety. We develop a novel automated verification framework for feed-forward multi-layer neural networks based on Satisfiability Modulo Theory (SMT). We focus on safety of image classification decisions with respect to image manipulations, such as scratches or changes to camera angle or lighting conditions that would result in the same class being assigned by a human, and define safety for an individual decision in terms of invariance of the classification within a small neighbourhood of the original image. We enable exhaustive search of the region by employing discretisation, and propagate the analysis layer by layer. Our method works directly with the network code and, in contrast to existing methods, can guarantee that adversarial examples, if they exist, are found for the given region and family of manipulations. If found, adversarial examples can be shown to human testers and/or used to fine-tune the network. We implement the techniques using Z3 and evaluate them on state-of-the-art networks, including regularised and deep learning networks. We also compare against existing techniques to search for adversarial examples and estimate network robustness.
translated by 谷歌翻译
Existing neural network verifiers compute a proof that each input is handled correctly under a given perturbation by propagating a symbolic abstraction of reachable values at each layer. This process is repeated from scratch independently for each input (e.g., image) and perturbation (e.g., rotation), leading to an expensive overall proof effort when handling an entire dataset. In this work, we introduce a new method for reducing this verification cost without losing precision based on a key insight that abstractions obtained at intermediate layers for different inputs and perturbations can overlap or contain each other. Leveraging our insight, we introduce the general concept of shared certificates, enabling proof effort reuse across multiple inputs to reduce overall verification costs. We perform an extensive experimental evaluation to demonstrate the effectiveness of shared certificates in reducing the verification cost on a range of datasets and attack specifications on image classifiers including the popular patch and geometric perturbations. We release our implementation at https://github.com/eth-sri/proof-sharing.
translated by 谷歌翻译
域特异性启发式方法是有效解决组合问题的必不可少的技术。当前将特定于域的启发式方法与答案集编程(ASP)集成的方法在处理基于部分分配的非单调指定的启发式方法时,这是不令人满意的。例如,在挑选尚未放入垃圾箱中的物品时,这种启发式方法经常发生。因此,我们介绍了ASP中域特异性启发式方法声明性规范的新颖语法和语义。我们的方法支持启发式陈述,依赖于解决过程中所维持的部分任务,这是不可能的。我们在Alpha中提供了一种实现,该实现使Alpha成为第一个支持声明指定的域特定启发式方法的懒惰的ASP系统。使用两个实际的示例域来证明我们的提议的好处。此外,我们使用我们的方法用A*实施知情},该搜索首次在ASP中解决。 A*应用于两个进一步的搜索问题。实验证实,结合懒惰的ASP解决方案和我们的新型启发式方法对于解决工业大小的问题至关重要。
translated by 谷歌翻译
我们介绍了一种新颖的方法来对计算机程序进行自动终止分析:我们使用神经网络来表示排名功能。排名函数映射程序状态为从下面界限并随着程序运行而减小的值;排名函数的存在证明了程序终止。我们从程序的采样执行轨迹训练神经网络,以使网络的输出沿轨迹降低;然后,我们使用符号推理正式验证其对所有可能执行的概括。通过肯定的答案,我们获得了该计划的正式终止证书,我们称之为神经排名函数。我们证明,由于神经网络代表非线性功能的能力,我们的方法成功地超过了最先进工具的程序。这包括在其循环条件和包括非线性表达式的程序中使用析取的程序。
translated by 谷歌翻译
Learning-enabled control systems have demonstrated impressive empirical performance on challenging control problems in robotics, but this performance comes at the cost of reduced transparency and lack of guarantees on the safety or stability of the learned controllers. In recent years, new techniques have emerged to provide these guarantees by learning certificates alongside control policies -- these certificates provide concise, data-driven proofs that guarantee the safety and stability of the learned control system. These methods not only allow the user to verify the safety of a learned controller but also provide supervision during training, allowing safety and stability requirements to influence the training process itself. In this paper, we provide a comprehensive survey of this rapidly developing field of certificate learning. We hope that this paper will serve as an accessible introduction to the theory and practice of certificate learning, both to those who wish to apply these tools to practical robotics problems and to those who wish to dive more deeply into the theory of learning for control.
translated by 谷歌翻译
结构分解方法,例如普遍的高树木分解,已成功用于解决约束满意度问题(CSP)。由于可以重复使用分解以求解具有相同约束范围的CSP,因此即使计算本身很难,将资源投资于计算良好的分解是有益的。不幸的是,即使示波器仅略有变化,当前方法也需要计算全新的分解。在本文中,我们迈出了解决CSP $ P $分解的问题的第一步,以使其成为由$ P $修改产生的新CSP $ P'$的有效分解。即使从理论上讲问题很难,我们还是提出并实施了一个有效更新GHD的框架。我们算法的实验评估强烈提出了实际适用性。
translated by 谷歌翻译
We present an approach for the verification of feed-forward neural networks in which all nodes have a piece-wise linear activation function. Such networks are often used in deep learning and have been shown to be hard to verify for modern satisfiability modulo theory (SMT) and integer linear programming (ILP) solvers.The starting point of our approach is the addition of a global linear approximation of the overall network behavior to the verification problem that helps with SMT-like reasoning over the network behavior. We present a specialized verification algorithm that employs this approximation in a search process in which it infers additional node phases for the non-linear nodes in the network from partial node phase assignments, similar to unit propagation in classical SAT solving. We also show how to infer additional conflict clauses and safe node fixtures from the results of the analysis steps performed during the search. The resulting approach is evaluated on collision avoidance and handwritten digit recognition case studies.
translated by 谷歌翻译
回答集编程(ASP)已成为一种流行的和相当复杂的声明问题解决方法。这是由于其具有吸引力的地址解决方案的工作流程,这是可以轻松解决问题解决的方法,即使对于计算机科学外的守护者而言。与此不同,底层技术的高度复杂性使得ASP专家越来越难以将想法付诸实践。有关解决此问题,本教程旨在使用户能够构建自己的基于ASP的系统。更确切地说,我们展示了ASP系统Clingo如何用于扩展ASP和实现定制的专用系统。为此,我们提出了两个替代方案。我们从传统的AI技术开始,并展示元编程如何用于扩展ASP。这是一种相当轻的方法,依赖于Clingo的reation特征来使用ASP本身表达新功能。与此不同,本教程的主要部分使用传统的编程(在Python中)来通过其应用程序编程接口操纵Clingo。这种方法允许改变和控制ASP的整个模型 - 地面解决工作流程。 COMENT of Clingo的新应用程序课程使我们能够通过自定义类似于Clingo中的进程来绘制Clingo的基础架构。例如,我们可能会互动到程序的抽象语法树,控制各种形式的多射击求解,并为外国推论设置理论传播者。另一种横截面结构,跨越元以及应用程序编程是Clingo的中间格式,即指定底层接地器和求解器之间的界面。我们通过示例和几个非琐碎的案例研究说明了本教程的前述概念和技术。
translated by 谷歌翻译