胰腺中的癌前囊肿或肿瘤的早期检测，即，导管内乳头状粘膜肿瘤（IPMN）是一项具有挑战性且复杂的任务，它可能导致更有利的结果。一旦检测到，还必须准确地对IPMN进行评分，因为低风险IPMN可以在监视计划下进行，而高危IPMN必须在变成癌症之前先手术切除。 IPMN分类的当前标准（Fukuoka等）显示出明显的操作员内和跨操作员变异性，除了容易出错，使适当的诊断不可靠。通过深度学习范式在人工智能方面的既定进展可能为有效支持胰腺癌的医疗决策提供了关键工具。在这项工作中，我们通过提出一种基于AI的新型IPMN分类器来遵循这一趋势，该分类器利用了Transformer网络最近在包括视觉的各种任务（包括视觉的任务）上概括的最新成功。我们特别表明，我们的基于变压器的模型比标准卷积神经网络更好地利用预训练，从而支持视觉中构建的构造统一性，包括医学图像域，并可以更好地解释获得的结果。

在医学领域，通常寻求多中心协作来通过利用患者和临床数据的异质性来产生更广泛的发现。但是，最近的隐私法规阻碍了共享数据的可能性，因此，提出了支持诊断和预后的基于机器学习的解决方案。联合学习（FL）旨在通过将基于AI的解决方案带入数据所有者，而仅共享需要汇总的本地AI模型或其部分，以避免这种限制。但是，大多数现有的联合学习解决方案仍处于起步阶段，并且由于缺乏可靠和有效的聚合计划能够保留本地学到的知识，从而显示出薄弱的隐私保护，因为可以从模型更新中重建实际数据，因此显示出几个缺点。此外，这些方法中的大多数，尤其是那些处理医学数据的方法，都依赖于一种集中的分布式学习策略，该策略构成了稳健性，可伸缩性和信任问题。在本文中，我们提出了一种分散的分布式方法，该方法从经验重播和生成对抗性研究中利用概念，有效地整合了本地节点的功能，从而提供了能够在维持隐私的同时跨多个数据集进行概括的模型。为了模拟现实的非i.i.d，使用多个数据集对两项任务进行了两项任务测试：结核病和黑色素瘤分类。数据方案。结果表明，我们的方法实现了与标准（未赋予）学习和联合方法相当的性能（因此，更有利）。

这项工作调查了持续学习（CL）与转移学习（TL）之间的纠缠。特别是，我们阐明了网络预训练的广泛应用，强调它本身受到灾难性遗忘的影响。不幸的是，这个问题导致在以后任务期间知识转移的解释不足。在此基础上，我们提出了转移而不忘记（TWF），这是在固定的经过预定的兄弟姐妹网络上建立的混合方法，该方法不断传播源域中固有的知识，通过层次损失项。我们的实验表明，TWF在各种设置上稳步优于其他CL方法，在各种数据集和不同的缓冲尺寸上，平均每种类型的精度增长了4.81％。

Recent years have seen a proliferation of research on adversarial machine learning. Numerous papers demonstrate powerful algorithmic attacks against a wide variety of machine learning (ML) models, and numerous other papers propose defenses that can withstand most attacks. However, abundant real-world evidence suggests that actual attackers use simple tactics to subvert ML-driven systems, and as a result security practitioners have not prioritized adversarial ML defenses. Motivated by the apparent gap between researchers and practitioners, this position paper aims to bridge the two domains. We first present three real-world case studies from which we can glean practical insights unknown or neglected in research. Next we analyze all adversarial ML papers recently published in top security conferences, highlighting positive trends and blind spots. Finally, we state positions on precise and cost-driven threat modeling, collaboration between industry and academia, and reproducible research. We believe that our positions, if adopted, will increase the real-world impact of future endeavours in adversarial ML, bringing both researchers and practitioners closer to their shared goal of improving the security of ML systems.

Deep spiking neural networks (SNNs) offer the promise of low-power artificial intelligence. However, training deep SNNs from scratch or converting deep artificial neural networks to SNNs without loss of performance has been a challenge. Here we propose an exact mapping from a network with Rectified Linear Units (ReLUs) to an SNN that fires exactly one spike per neuron. For our constructive proof, we assume that an arbitrary multi-layer ReLU network with or without convolutional layers, batch normalization and max pooling layers was trained to high performance on some training set. Furthermore, we assume that we have access to a representative example of input data used during training and to the exact parameters (weights and biases) of the trained ReLU network. The mapping from deep ReLU networks to SNNs causes zero percent drop in accuracy on CIFAR10, CIFAR100 and the ImageNet-like data sets Places365 and PASS. More generally our work shows that an arbitrary deep ReLU network can be replaced by an energy-efficient single-spike neural network without any loss of performance.

Deep learning-based object detection is a powerful approach for detecting faulty insulators in power lines. This involves training an object detection model from scratch, or fine tuning a model that is pre-trained on benchmark computer vision datasets. This approach works well with a large number of insulator images, but can result in unreliable models in the low data regime. The current literature mainly focuses on detecting the presence or absence of insulator caps, which is a relatively easy detection task, and does not consider detection of finer faults such as flashed and broken disks. In this article, we formulate three object detection tasks for insulator and asset inspection from aerial images, focusing on incipient faults in disks. We curate a large reference dataset of insulator images that can be used to learn robust features for detecting healthy and faulty insulators. We study the advantage of using this dataset in the low target data regime by pre-training on the reference dataset followed by fine-tuning on the target dataset. The results suggest that object detection models can be used to detect faults in insulators at a much incipient stage, and that transfer learning adds value depending on the type of object detection model. We identify key factors that dictate performance in the low data-regime and outline potential approaches to improve the state-of-the-art.

Deploying machine learning models in production may allow adversaries to infer sensitive information about training data. There is a vast literature analyzing different types of inference risks, ranging from membership inference to reconstruction attacks. Inspired by the success of games (i.e., probabilistic experiments) to study security properties in cryptography, some authors describe privacy inference risks in machine learning using a similar game-based style. However, adversary capabilities and goals are often stated in subtly different ways from one presentation to the other, which makes it hard to relate and compose results. In this paper, we present a game-based framework to systematize the body of knowledge on privacy inference risks in machine learning.

A systematic review on machine-learning strategies for improving generalizability (cross-subjects and cross-sessions) electroencephalography (EEG) based in emotion classification was realized. In this context, the non-stationarity of EEG signals is a critical issue and can lead to the Dataset Shift problem. Several architectures and methods have been proposed to address this issue, mainly based on transfer learning methods. 418 papers were retrieved from the Scopus, IEEE Xplore and PubMed databases through a search query focusing on modern machine learning techniques for generalization in EEG-based emotion assessment. Among these papers, 75 were found eligible based on their relevance to the problem. Studies lacking a specific cross-subject and cross-session validation strategy and making use of other biosignals as support were excluded. On the basis of the selected papers' analysis, a taxonomy of the studies employing Machine Learning (ML) methods was proposed, together with a brief discussion on the different ML approaches involved. The studies with the best results in terms of average classification accuracy were identified, supporting that transfer learning methods seem to perform better than other approaches. A discussion is proposed on the impact of (i) the emotion theoretical models and (ii) psychological screening of the experimental sample on the classifier performances.

We extend best-subset selection to linear Multi-Task Learning (MTL), where a set of linear models are jointly trained on a collection of datasets (``tasks''). Allowing the regression coefficients of tasks to have different sparsity patterns (i.e., different supports), we propose a modeling framework for MTL that encourages models to share information across tasks, for a given covariate, through separately 1) shrinking the coefficient supports together, and/or 2) shrinking the coefficient values together. This allows models to borrow strength during variable selection even when the coefficient values differ markedly between tasks. We express our modeling framework as a Mixed-Integer Program, and propose efficient and scalable algorithms based on block coordinate descent and combinatorial local search. We show our estimator achieves statistically optimal prediction rates. Importantly, our theory characterizes how our estimator leverages the shared support information across tasks to achieve better variable selection performance. We evaluate the performance of our method in simulations and two biology applications. Our proposed approaches outperform other sparse MTL methods in variable selection and prediction accuracy. Interestingly, penalties that shrink the supports together often outperform penalties that shrink the coefficient values together. We will release an R package implementing our methods.

This thesis develops the translation between category theory and computational linguistics as a foundation for natural language processing. The three chapters deal with syntax, semantics and pragmatics. First, string diagrams provide a unified model of syntactic structures in formal grammars. Second, functors compute semantics by turning diagrams into logical, tensor, neural or quantum computation. Third, the resulting functorial models can be composed to form games where equilibria are the solutions of language processing tasks. This framework is implemented as part of DisCoPy, the Python library for computing with string diagrams. We describe the correspondence between categorical, linguistic and computational structures, and demonstrate their applications in compositional natural language processing.