The Internet of Things (IoT) is a system that connects physical computing devices, sensors, software, and other technologies. Data can be collected, transferred, and exchanged with other devices over the network without requiring human interactions. One challenge the development of IoT faces is the existence of anomaly data in the network. Therefore, research on anomaly detection in the IoT environment has become popular and necessary in recent years. This survey provides an overview to understand the current progress of the different anomaly detection algorithms and how they can be applied in the context of the Internet of Things. In this survey, we categorize the widely used anomaly detection machine learning and deep learning techniques in IoT into three types: clustering-based, classification-based, and deep learning based. For each category, we introduce some state-of-the-art anomaly detection methods and evaluate the advantages and limitations of each technique.

In 2021 300 mm of rain, nearly half the average annual rainfall, fell near Catania (Sicily island, Italy). Such events took place in just a few hours, with dramatic consequences on the environmental, social, economic, and health systems of the region. This is the reason why, detecting extreme rainfall events is a crucial prerequisite for planning actions able to reverse possibly intensified dramatic future scenarios. In this paper, the Affinity Propagation algorithm, a clustering algorithm grounded on machine learning, was applied, to the best of our knowledge, for the first time, to identify excess rain events in Sicily. This was possible by using a high-frequency, large dataset we collected, ranging from 2009 to 2021 which we named RSE (the Rainfall Sicily Extreme dataset). Weather indicators were then been employed to validate the results, thus confirming the presence of recent anomalous rainfall events in eastern Sicily. We believe that easy-to-use and multi-modal data science techniques, such as the one proposed in this study, could give rise to significant improvements in policy-making for successfully contrasting climate changes.

与原子分辨率上可实现的分子量相比，粗晶片（CG）能够研究较大系统和更长的时间尺度的分子特性。最近已经提出了机器学习技术来学习CG粒子相互作用，即开发CG力场。分子的图表和图形卷积神经网络结构的监督训练用于通过力匹配方案来学习平均力的潜力。在这项工作中，作用在每个CG粒子上的力与以Schnet的名义相关的其本地环境的表示，该代表通过连续过滤器卷积构建。我们探讨了Schnet模型在获得液体苯的CG潜力的应用，研究模型结构和超参数对模拟CG系统的热力学，动力学和结构特性的影响，并报告和讨论所设想的挑战以及未来的指导。

由于大分子系统中存在的各种时间尺度，其计算研究是必要的。粗粒（CG）允许在不同的系统分辨率之间建立联系，并为开发强大的多尺度模拟和分析提供骨干。 CG映射过程通常是系统和特定于应用程序的，它依赖于化学直觉。在这项工作中，我们探讨了基于变异自动编码器的机器学习策略的应用，以开发合适的映射方案，从原子体到分子的粗粒空间，并随着化学复杂性的增加而开发。对模型超级法对训练过程和最终输出的影响进行了广泛的评估，并通过定义不同的损失函数的定义进行了现有方法，并实施了确保输出物理一致性的选择标准。分析了输入特征选择与重建精度之间的关系，从而支持将旋转不变性引入系统的需求。在映射和背景步骤中，该方法的优势和局限性都得到了强调和严格的讨论。

Hawkes流程最近从机器学习社区中引起了人们对建模事件序列数据的多功能性的越来越多的关注。尽管它们具有丰富的历史可以追溯到几十年前，但其某些属性（例如用于学习参数的样本复杂性和释放差异化私有版本的样本复杂性）尚未得到彻底的分析。在这项工作中，我们研究了具有背景强度$ \ mu $和激发功能$ \ alpha e^{ - \ beta t} $的标准霍克斯进程。我们提供$ \ mu $和$ \ alpha $的非私人和差异私人估计器，并在两种设置中获得样本复杂性结果以量化隐私成本。我们的分析利用了霍克斯过程的强大混合特性和经典的中央限制定理的结果，结果较弱的随机变量。我们在合成数据集和真实数据集上验证了我们的理论发现。

在基于预测的决策系统中，不同的观点可能是矛盾的：决策者的短期业务目标通常与决策主体的愿望相抵触。平衡这两个观点是一个价值问题。我们提供一个框架，使这些具有价值的选择清晰可见。为此，我们假设我们获得了一个训练有素的模型，并希望找到平衡决策者和决策主体观点的决策规则。我们提供了一种形式化这两种观点的方法，即评估决策者的效用和对决策主体的公平性。在这两种情况下，这个想法都是从决策者和决策主题中引起的价值观，然后将其变成可衡量的东西。为了进行公平评估，我们以基于福利的公平性的文献为基础，并询问公用事业（或福利）的公平分布是什么样的。在此步骤中，我们以分配正义的著名理论为基础。这使我们能够得出一个公平分数，然后将其与许多不同决策规则的决策者实用程序进行比较。这样，我们提供了一种平衡决策者的实用性的方法，以及对基于预测的决策系统的决策主体的公平性。

Traditionally, deep learning methods for breast cancer classification perform a single-view analysis. However, radiologists simultaneously analyze all four views that compose a mammography exam, owing to the correlations contained in mammography views, which present crucial information for identifying tumors. In light of this, some studies have started to propose multi-view methods. Nevertheless, in such existing architectures, mammogram views are processed as independent images by separate convolutional branches, thus losing correlations among them. To overcome such limitations, in this paper we propose a novel approach for multi-view breast cancer classification based on parameterized hypercomplex neural networks. Thanks to hypercomplex algebra properties, our networks are able to model, and thus leverage, existing correlations between the different views that comprise a mammogram, thus mimicking the reading process performed by clinicians. The proposed methods are able to handle the information of a patient altogether without breaking the multi-view nature of the exam. We define architectures designed to process two-view exams, namely PHResNets, and four-view exams, i.e., PHYSEnet and PHYBOnet. Through an extensive experimental evaluation conducted with publicly available datasets, we demonstrate that our proposed models clearly outperform real-valued counterparts and also state-of-the-art methods, proving that breast cancer classification benefits from the proposed multi-view architectures. We also assess the method's robustness beyond mammogram analysis by considering different benchmarks, as well as a finer-scaled task such as segmentation. Full code and pretrained models for complete reproducibility of our experiments are freely available at: https://github.com/ispamm/PHBreast.

Spatial audio methods are gaining a growing interest due to the spread of immersive audio experiences and applications, such as virtual and augmented reality. For these purposes, 3D audio signals are often acquired through arrays of Ambisonics microphones, each comprising four capsules that decompose the sound field in spherical harmonics. In this paper, we propose a dual quaternion representation of the spatial sound field acquired through an array of two First Order Ambisonics (FOA) microphones. The audio signals are encapsulated in a dual quaternion that leverages quaternion algebra properties to exploit correlations among them. This augmented representation with 6 degrees of freedom (6DOF) involves a more accurate coverage of the sound field, resulting in a more precise sound localization and a more immersive audio experience. We evaluate our approach on a sound event localization and detection (SELD) benchmark. We show that our dual quaternion SELD model with temporal convolution blocks (DualQSELD-TCN) achieves better results with respect to real and quaternion-valued baselines thanks to our augmented representation of the sound field. Full code is available at: https://github.com/ispamm/DualQSELD-TCN.

从Chaser Spacecraft发射的系绳网提供了有希望的方法，可以在轨道中捕获和处理大型空间碎片。该系绳网络系统受到影响和致动的几种不确定性来源，影响其净爆发和关闭控制的性能。然而，设计控制动作的早期可靠性的优化方法仍然具有挑战性，并计算到相对于追逐者相对于追逐者的不同发射方案和目标（碎片）状态概括。为了搜索一般和可靠的控制策略，本文介绍了一种加强学习框架，它集成了具有净动力学模拟的近端策略优化（PPO2）方法。后者允许评估基于网络的目标捕获的剧集，并估算捕获质量索引，作为PPO2的奖励反馈。在这里，在任何给定的发射方案下，学习的策略旨在根据移动网和目标的状态来模拟网络结束动作的定时。考虑了随机状态转换模型，以便在国家估算和发射致动中纳入合成不确定性。随着培训期间的显着奖励改进，训练有素的策略表明捕获性能（在广泛的发射/目标场景范围内），接近基于可靠性的优化在各个方案上运行。

事实证明，超复杂的神经网络可以减少参数的总数，同时通过利用Clifford代数的特性来确保有价值的性能。最近，通过涉及有效的参数化kronecker产品，超复合线性层得到了进一步改善。在本文中，我们定义了超复杂卷积层的参数化，并介绍了轻巧有效的大型大型模型的参数化超复杂神经网络（PHNN）。我们的方法直接从数据中掌握了卷积规则和过滤器组织，而无需遵循严格的预定义域结构。 Phnns可以灵活地在任何用户定义或调谐域中操作，无论代数规则是否是预设的，从1D到$ n $ d。这样的锻造性允许在其自然域中处理多维输入，而无需吞并进一步的尺寸，而是在Quaternion神经网络中使用3D输入（例如颜色图像）。结果，拟议中的Phnn家族以$ 1/n $的参数运行，因为其在真实域中的类似物。我们通过在各种图像数据集上执行实验以及音频数据集证明了这种方法对应用程序多个域的多功能性，在这些实验中，我们的方法的表现优于真实和Quaternion值值。完整代码可在以下网址获得：https：//github.com/elegan23/hypernets。