Federated learning is a learning paradigm to enable collaborative learning across different parties without revealing raw data. Notably, vertical federated learning (VFL), where parties share the same set of samples but only hold partial features, has a wide range of real-world applications. However, most existing studies in VFL disregard the "record linkage" process. They design algorithms either assuming the data from different parties can be exactly linked or simply linking each record with its most similar neighboring record. These approaches may fail to capture the key features from other less similar records. Moreover, such improper linkage cannot be corrected by training since existing approaches provide no feedback on linkage during training. In this paper, we design a novel coupled training paradigm, FedSim, that integrates one-to-many linkage into the training process. Besides enabling VFL in many real-world applications with fuzzy identifiers, FedSim also achieves better performance in traditional VFL tasks. Moreover, we theoretically analyze the additional privacy risk incurred by sharing similarities. Our experiments on eight datasets with various similarity metrics show that FedSim outperforms other state-of-the-art baselines. The codes of FedSim are available at https://github.com/Xtra-Computing/FedSim.

The presence of offensive language on social media platforms and the implications this poses is becoming a major concern in modern society. Given the enormous amount of content created every day, automatic methods are required to detect and deal with this type of content. Until now, most of the research has focused on solving the problem for the English language, while the problem is multilingual. We construct a Danish dataset containing user-generated comments from \textit{Reddit} and \textit{Facebook}. It contains user generated comments from various social media platforms, and to our knowledge, it is the first of its kind. Our dataset is annotated to capture various types and target of offensive language. We develop four automatic classification systems, each designed to work for both the English and the Danish language. In the detection of offensive language in English, the best performing system achieves a macro averaged F1-score of $0.74$, and the best performing system for Danish achieves a macro averaged F1-score of $0.70$. In the detection of whether or not an offensive post is targeted, the best performing system for English achieves a macro averaged F1-score of $0.62$, while the best performing system for Danish achieves a macro averaged F1-score of $0.73$. Finally, in the detection of the target type in a targeted offensive post, the best performing system for English achieves a macro averaged F1-score of $0.56$, and the best performing system for Danish achieves a macro averaged F1-score of $0.63$. Our work for both the English and the Danish language captures the type and targets of offensive language, and present automatic methods for detecting different kinds of offensive language such as hate speech and cyberbullying.

This paper focuses on the application of one type of empirical methods, namely statistical causal inference (SCI, see section 2). Such methods have their roots in a number of applied fields (from AI to econometrics) and aim to provide a framework for making valid inferences about causal effects based on interventional or observational data. More specifically, we focus on SCI methods that use graphical models as developed by Pearl and colleagues [1, 2]. This framework has been shown to be equivalent of the potential-outcomes framework (also called the Neyman-Rubin Causal Model [3]) but enriches it by making use of an explicit causal structure called a graphical causal model. Making assumptions about causal effects explicit through a graphical structure has several advantages. First, it helps determine whether causal effects can be estimated and how they might be estimated (see section 2).

As neural networks increasingly make critical decisions in high-stakes settings, monitoring and explaining their behavior in an understandable and trustworthy manner is a necessity. One commonly used type of explainer is post hoc feature attribution, a family of methods for giving each feature in an input a score corresponding to its influence on a model's output. A major limitation of this family of explainers in practice is that they can disagree on which features are more important than others. Our contribution in this paper is a method of training models with this disagreement problem in mind. We do this by introducing a Post hoc Explainer Agreement Regularization (PEAR) loss term alongside the standard term corresponding to accuracy, an additional term that measures the difference in feature attribution between a pair of explainers. We observe on three datasets that we can train a model with this loss term to improve explanation consensus on unseen data, and see improved consensus between explainers other than those used in the loss term. We examine the trade-off between improved consensus and model performance. And finally, we study the influence our method has on feature attribution explanations.

Many methods that build powerful variational distributions based on unadjusted Langevin transitions exist. Most of these were developed using a wide range of different approaches and techniques. Unfortunately, the lack of a unified analysis and derivation makes developing new methods and reasoning about existing ones a challenging task. We address this giving a single analysis that unifies and generalizes these existing techniques. The main idea is to augment the target and variational by numerically simulating the underdamped Langevin diffusion process and its time reversal. The benefits of this approach are twofold: it provides a unified formulation for many existing methods, and it simplifies the development of new ones. In fact, using our formulation we propose a new method that combines the strengths of previously existing algorithms; it uses underdamped Langevin transitions and powerful augmentations parameterized by a score network. Our empirical evaluation shows that our proposed method consistently outperforms relevant baselines in a wide range of tasks.

As AI-based decision systems proliferate, their successful operationalization requires balancing multiple desiderata: predictive performance, disparity across groups, safeguarding sensitive group attributes (e.g., race), and engineering cost. We present a holistic framework for evaluating and contextualizing fairness interventions with respect to the above desiderata. The two key points of practical consideration are \emph{where} (pre-, in-, post-processing) and \emph{how} (in what way the sensitive group data is used) the intervention is introduced. We demonstrate our framework with a case study on predictive parity. In it, we first propose a novel method for achieving predictive parity fairness without using group data at inference time via distibutionally robust optimization. Then, we showcase the effectiveness of these methods in a benchmarking study of close to 400 variations across two major model types (XGBoost vs. Neural Net), ten datasets, and over twenty unique methodologies. Methodological insights derived from our empirical study inform the practical design of ML workflow with fairness as a central concern. We find predictive parity is difficult to achieve without using group data, and despite requiring group data during model training (but not inference), distributionally robust methods we develop provide significant Pareto improvement. Moreover, a plain XGBoost model often Pareto-dominates neural networks with fairness interventions, highlighting the importance of model inductive bias.

Automatic language identification is a challenging problem. Discriminating between closely related languages is especially difficult. This paper presents a machine learning approach for automatic language identification for the Nordic languages, which often suffer miscategorisation by existing state-of-the-art tools. Concretely we will focus on discrimination between six Nordic languages: Danish, Swedish, Norwegian (Nynorsk), Norwegian (Bokm{\aa}l), Faroese and Icelandic.

Neurophysiological time series recordings like the electroencephalogram (EEG) or local field potentials are obtained from multiple sensors. They can be decoded by machine learning models in order to estimate the ongoing brain state of a patient or healthy user. In a brain-computer interface (BCI), this decoded brain state information can be used with minimal time delay to either control an application, e.g., for communication or for rehabilitation after stroke, or to passively monitor the ongoing brain state of the subject, e.g., in a demanding work environment. A specific challenge in such decoding tasks is posed by the small dataset sizes in BCI compared to other domains of machine learning like computer vision or natural language processing. A possibility to tackle classification or regression problems in BCI despite small training data sets is through transfer learning, which utilizes data from other sessions, subjects or even datasets to train a model. In this exploratory study, we propose novel domain-specific embeddings for neurophysiological data. Our approach is based on metric learning and builds upon the recently proposed ladder loss. Using embeddings allowed us to benefit, both from the good generalisation abilities and robustness of deep learning and from the fast training of classical machine learning models for subject-specific calibration. In offline analyses using EEG data of 14 subjects, we tested the embeddings' feasibility and compared their efficiency with state-of-the-art deep learning models and conventional machine learning pipelines. In summary, we propose the use of metric learning to obtain pre-trained embeddings of EEG-BCI data as a means to incorporate domain knowledge and to reach competitive performance on novel subjects with minimal calibration requirements.

We investigate the convergence and convergence rate of stochastic training algorithms for Neural Networks (NNs) that have been inspired by Dropout (Hinton et al., 2012). With the goal of avoiding overfitting during training of NNs, dropout algorithms consist in practice of multiplying the weight matrices of a NN componentwise by independently drawn random matrices with $\{0, 1 \}$-valued entries during each iteration of Stochastic Gradient Descent (SGD). This paper presents a probability theoretical proof that for fully-connected NNs with differentiable, polynomially bounded activation functions, if we project the weights onto a compact set when using a dropout algorithm, then the weights of the NN converge to a unique stationary point of a projected system of Ordinary Differential Equations (ODEs). After this general convergence guarantee, we go on to investigate the convergence rate of dropout. Firstly, we obtain generic sample complexity bounds for finding $\epsilon$-stationary points of smooth nonconvex functions using SGD with dropout that explicitly depend on the dropout probability. Secondly, we obtain an upper bound on the rate of convergence of Gradient Descent (GD) on the limiting ODEs of dropout algorithms for NNs with the shape of arborescences of arbitrary depth and with linear activation functions. The latter bound shows that for an algorithm such as Dropout or Dropconnect (Wan et al., 2013), the convergence rate can be impaired exponentially by the depth of the arborescence. In contrast, we experimentally observe no such dependence for wide NNs with just a few dropout layers. We also provide a heuristic argument for this observation. Our results suggest that there is a change of scale of the effect of the dropout probability in the convergence rate that depends on the relative size of the width of the NN compared to its depth.

Recent studies showed that Photoplethysmography (PPG) sensors embedded in wearable devices can estimate heart rate (HR) with high accuracy. However, despite of prior research efforts, applying PPG sensor based HR estimation to embedded devices still faces challenges due to the energy-intensive high-frequency PPG sampling and the resource-intensive machine-learning models. In this work, we aim to explore HR estimation techniques that are more suitable for lower-power and resource-constrained embedded devices. More specifically, we seek to design techniques that could provide high-accuracy HR estimation with low-frequency PPG sampling, small model size, and fast inference time. First, we show that by combining signal processing and ML, it is possible to reduce the PPG sampling frequency from 125 Hz to only 25 Hz while providing higher HR estimation accuracy. This combination also helps to reduce the ML model feature size, leading to smaller models. Additionally, we present a comprehensive analysis on different ML models and feature sizes to compare their accuracy, model size, and inference time. The models explored include Decision Tree (DT), Random Forest (RF), K-nearest neighbor (KNN), Support vector machines (SVM), and Multi-layer perceptron (MLP). Experiments were conducted using both a widely-utilized dataset and our self-collected dataset. The experimental results show that our method by combining signal processing and ML had only 5% error for HR estimation using low-frequency PPG data. Moreover, our analysis showed that DT models with 10 to 20 input features usually have good accuracy, while are several magnitude smaller in model sizes and faster in inference time.