The training of classification models for fault diagnosis tasks using geographically dispersed data is a crucial task for original equipment manufacturers (OEMs) seeking to provide long-term service contracts (LTSCs) to their customers. Due to privacy and bandwidth constraints, such models must be trained in a federated fashion. Moreover, due to harsh industrial settings the data often suffers from feature and label uncertainty. Therefore, we study the problem of training a distributionally robust (DR) support vector machine (SVM) in a federated fashion over a network comprised of a central server and $G$ clients without sharing data. We consider the setting where the local data of each client $g$ is sampled from a unique true distribution $\mathbb{P}_g$, and the clients can only communicate with the central server. We propose a novel Mixture of Wasserstein Balls (MoWB) ambiguity set that relies on local Wasserstein balls centered at the empirical distribution of the data at each client. We study theoretical aspects of the proposed ambiguity set, deriving its out-of-sample performance guarantees and demonstrating that it naturally allows for the separability of the DR problem. Subsequently, we propose two distributed optimization algorithms for training the global FDR-SVM: i) a subgradient method-based algorithm, and ii) an alternating direction method of multipliers (ADMM)-based algorithm. We derive the optimization problems to be solved by each client and provide closed-form expressions for the computations performed by the central server during each iteration for both algorithms. Finally, we thoroughly examine the performance of the proposed algorithms in a series of numerical experiments utilizing both simulation data and popular real-world datasets.

In applications, an anticipated situation is where the system of interest has never been encountered before and sparse observations can be made only once. Can the dynamics be faithfully reconstructed from the limited observations without any training data? This problem defies any known traditional methods of nonlinear time-series analysis as well as existing machine-learning methods that typically require extensive data from the target system for training. We address this challenge by developing a hybrid transformer and reservoir-computing machine-learning scheme. The key idea is that, for a complex and nonlinear target system, the training of the transformer can be conducted not using any data from the target system, but with essentially unlimited synthetic data from known chaotic systems. The trained transformer is then tested with the sparse data from the target system. The output of the transformer is further fed into a reservoir computer for predicting the long-term dynamics or the attractor of the target system. The power of the proposed hybrid machine-learning framework is demonstrated using a large number of prototypical nonlinear dynamical systems, with high reconstruction accuracy even when the available data is only 20% of that required to faithfully represent the dynamical behavior of the underlying system. The framework provides a paradigm of reconstructing complex and nonlinear dynamics in the extreme situation where training data does not exist and the observations are random and sparse.

The intriguing in-context learning (ICL) abilities of deep Transformer models have lately garnered significant attention. By studying in-context linear regression on unimodal Gaussian data, recent empirical and theoretical works have argued that ICL emerges from Transformers' abilities to simulate learning algorithms like gradient descent. However, these works fail to capture the remarkable ability of Transformers to learn multiple tasks in context. To this end, we study in-context learning for linear regression with diverse tasks, characterized by data covariance matrices with condition numbers ranging from $[1, \kappa]$, and highlight the importance of depth in this setting. More specifically, (a) we show theoretical lower bounds of $\log(\kappa)$ (or $\sqrt{\kappa}$) linear attention layers in the unrestricted (or restricted) attention setting and, (b) we show that multilayer Transformers can indeed solve such tasks with a number of layers that matches the lower bounds. However, we show that this expressivity of multilayer Transformer comes at the price of robustness. In particular, multilayer Transformers are not robust to even distributional shifts as small as $O(e^{-L})$ in Wasserstein distance, where $L$ is the depth of the network. We then demonstrate that Looped Transformers -- a special class of multilayer Transformers with weight-sharing -- not only exhibit similar expressive power but are also provably robust under mild assumptions. Besides out-of-distribution generalization, we also show that Looped Transformers are the only models that exhibit a monotonic behavior of loss with respect to depth.

The library is based on distributionally robust optimization using optimal transport distances. For ease of use, it features both scikit-learn compatible estimators for popular objectives, as well as a wrapper for PyTorch modules, enabling researchers and practitioners to use it in a wide range of models with minimal code changes. Its implementation relies on an entropic smoothing of the original robust objective in order to ensure maximal model flexibility. The library is available at https://github.com/iutzeler/skwdro. Keywords: Distributionally robust optim., distribution shifts, entropic regularization

The use of graph-based models had already been a key element in applications such as route planning (e.g., Dijkstra's algorithm) [3], community detection (e.g., clique percolation or Louvain algorithms) [4] or the analysis of complex networks such as biological and social networks [5, 6]. The representation of manifolds by means of graphs in the field of semi-supervised learning is another example of graph-powered applications [7]. Overall, the GSP framework [8] has enabled the use and development of novel techniques on data residing in graphs, thus emerging as an alternative to classical machine learning techniques that do not make explicit use of data structure. In this way, the graph topology, which represents the relationships between the graph's nodes, is fed to graph-based models that explicitly model the structure of the data [9]. A wide variety of concepts have been applied to signals defined over graphs, such as signal shift, translation, convolution, or filtering [10]. An important concept of the GSP is the notion of signal smoothness, also expressed via the total variation (TV) or the Dirichlet energy, which are quadratic forms and can be used to evaluate how a signal fits a given graph structure or vice-versa [11]. This idea is linked with the Graph Discrete Fourier Transform (GDFT) that makes use of the graph topology to obtain the graph Fourier basis and allow the computation of the transform coefficients of a graph signal and also led to the development of graph filters [12]. In the field of machine learning, graphs have been used as regularizers in optimization problems, e.g., the regularization of neural networks for semi-supervised learning tasks [13].

Foundation Language Models (FLMs) such as BERT and its variants have achieved remarkable success in natural language processing. To date, the interpretability of FLMs has primarily relied on the attention weights in their self-attention layers. However, these attention weights only provide word-level interpretations, failing to capture higher-level structures, and are therefore lacking in readability and intuitiveness. To address this challenge, we first provide a formal definition of conceptual interpretation and then propose a variational Bayesian framework, dubbed VAriational Language Concept (VALC), to go beyond word-level interpretations and provide concept-level interpretations. Our theoretical analysis shows that our VALC finds the optimal language concepts to interpret FLM predictions. Empirical results on several real-world datasets show that our method can successfully provide conceptual interpretation for FLMs.

Generating synthetic tabular data is critical in machine learning, especially when real data is limited or sensitive. Traditional generative models often face challenges due to the unique characteristics of tabular data, such as mixed data types and varied distributions, and require complex preprocessing or large pretrained models. In this paper, we introduce a novel, lossless binary transformation method that converts any tabular data into fixed-size binary representations, and a corresponding new generative model called Binary Diffusion, specifically designed for binary data. Binary Diffusion leverages the simplicity of XOR operations for noise addition and removal and employs binary cross-entropy loss for training. Our approach eliminates the need for extensive preprocessing, complex noise parameter tuning, and pretraining on large datasets. We evaluate our model on several popular tabular benchmark datasets, demonstrating that Binary Diffusion outperforms existing state-of-the-art models on Travel, Adult Income, and Diabetes datasets while being significantly smaller in size.

Federated learning is renowned for its efficacy in distributed model training, ensuring that users, called clients, retain data privacy by not disclosing their data to the central server that orchestrates collaborations. Most previous work on federated learning assumes that clients possess static batches of training data. However, clients may also need to make real-time predictions on streaming data in non-stationary environments. In such dynamic environments, employing pre-trained models may be inefficient, as they struggle to adapt to the constantly evolving data streams. To address this challenge, clients can fine-tune models online, leveraging their observed data to enhance performance. Despite the potential benefits of client participation in federated online model fine-tuning, existing analyses have not conclusively demonstrated its superiority over local model fine-tuning. To bridge this gap, the present paper develops a novel personalized federated learning algorithm, wherein each client constructs a personalized model by combining a locally fine-tuned model with multiple federated models learned by the server over time. Theoretical analysis and experiments on real datasets corroborate the effectiveness of this approach for real-time predictions and federated model fine-tuning.

There has been a surge of interest in leveraging speech as a marker of health for a wide spectrum of conditions. The underlying premise is that any neurological, mental, or physical deficits that impact speech production can be objectively assessed via automated analysis of speech. Recent advances in speech-based Artificial Intelligence (AI) models for diagnosing and tracking mental health, cognitive, and motor disorders often use supervised learning, similar to mainstream speech technologies like recognition and verification. However, clinical speech AI has distinct challenges, including the need for specific elicitation tasks, small available datasets, diverse speech representations, and uncertain diagnostic labels. As a result, application of the standard supervised learning paradigm may lead to models that perform well in controlled settings but fail to generalize in real-world clinical deployments. With translation into real-world clinical scenarios in mind, this tutorial paper provides an overview of the key components required for robust development of clinical speech AI. Specifically, this paper will cover the design of speech elicitation tasks and protocols most appropriate for different clinical conditions, collection of data and verification of hardware, development and validation of speech representations designed to measure clinical constructs of interest, development of reliable and robust clinical prediction models, and ethical and participant considerations for clinical speech AI. The goal is to provide comprehensive guidance on building models whose inputs and outputs link to the more interpretable and clinically meaningful aspects of speech, that can be interrogated and clinically validated on clinical datasets, and that adhere to ethical, privacy, and security considerations by design.

Distributed learning is essential to train machine learning algorithms across heterogeneous agents while maintaining data privacy. We conduct an asymptotic analysis of Unified Distributed SGD (UD-SGD), exploring a variety of communication patterns, including decentralized SGD and local SGD within Federated Learning (FL), as well as the increasing communication interval in the FL setting. In this study, we assess how different sampling strategies, such as i.i.d. sampling, shuffling, and Markovian sampling, affect the convergence speed of UD-SGD by considering the impact of agent dynamics on the limiting covariance matrix as described in the Central Limit Theorem (CLT). Our findings not only support existing theories on linear speedup and asymptotic network independence, but also theoretically and empirically show how efficient sampling strategies employed by individual agents contribute to overall convergence in UD-SGD. Simulations reveal that a few agents using highly efficient sampling can achieve or surpass the performance of the majority employing moderately improved strategies, providing new insights beyond traditional analyses focusing on the worst-performing agent.