While a number of studies have investigated speech and language features for the detection of AD and mild cognitive impairment (Fraser et al., 2016), and proposed various signal processing and machine learning methods for this task (Petti et al., 2020), the field still lacks balanced benchmark data against which different approaches can be systematically compared. This Research Topic addresses this issue by exploring the use of speech characteristics for AD recognition using balanced data and shared tasks, such as those provided by the ADReSS Challenges (Luz et al., 2020(Luz et al., , 2021. These tasks have brought together groups working on this active area of research, providing the community with benchmarks for comparison of speech and language approaches to cognitive assessment. Reflecting the multidisciplinary character of the topic, the articles in this collection span three journals: Frontiers of Aging Neuroscience, Frontiers of Computer Science and Frontiers in Psychology.Most papers in this Reseach Topic target two main tasks: AD classification, for distinguishing individuals with AD from healthy controls, and cognitive test score regression, to infer the patient's Mini Mental Status Examination (MMSE) score (Folstein et al., 1975). Of the twenty papers published in this collection, 14 used the ADReSS dataset (Luz et al., 2020), by itself or in combination with other data. The ADReSS dataset is a curated subset of DementiaBank's Pitt Corpus, matched for age and ge...

Attention based models such as Transformers involve pairwise interactions between data points, modeled with a learnable attention matrix. Importantly, this attention matrix is normalized with the SoftMax operator, which makes it row-wise stochastic. In this paper, we propose instead to use Sinkhorn's algorithm to make attention matrices doubly stochastic. We call the resulting model a Sinkformer. We show that the row-wise stochastic attention matrices in classical Transformers get close to doubly stochastic matrices as the number of epochs increases, justifying the use of Sinkhorn normalization as an informative prior. On the theoretical side, we show that, unlike the SoftMax operation, this normalization makes it possible to understand the iterations of self-attention modules as a discretized gradient-flow for the Wasserstein metric. We also show in the infinite number of samples limit that, when rescaling both attention matrices and depth, Sinkformers operate a heat diffusion. On the experimental side, we show that Sinkformers enhance model accuracy in vision and natural language processing tasks. In particular, on 3D shapes classification, Sinkformers lead to a significant improvement.

Over the past few years, several adversarial training methods have been proposed to improve the robustness of machine learning models against adversarial perturbations in the input. Despite remarkable progress in this regard, adversarial training is often observed to drop the standard test accuracy. This phenomenon has intrigued the research community to investigate the potential tradeoff between standard and robust accuracy as two performance measures. In this paper, we revisit this tradeoff for latent models and argue that this tradeoff is mitigated when the data enjoys a low-dimensional structure. In particular, we consider binary classification under two data generative models, namely Gaussian mixture model and generalized linear model, where the feature data lie on a low-dimensional manifold. We show that as the manifold dimension to the ambient dimension decreases, one can obtain models that are nearly optimal with respect to both, the standard accuracy and the robust accuracy measures.

In recent decades, biomedical signals have been used for communication in Human-Computer Interfaces (HCI) for medical applications; an instance of these signals are the myoelectric signals (MES), which are generated in the muscles of the human body as unidimensional patterns. Because of this, the methods and algorithms developed for pattern recognition in signals can be applied for their analyses once these signals have been sampled and turned into electromyographic (EMG) signals. Additionally, in recent years, many researchers have dedicated their efforts to studying prosthetic control utilizing EMG signal classification, that is, by logging a set of MES in a proper range of frequencies to classify the corresponding EMG signals. The feature classification can be carried out on the time domain or by using other domains such as the frequency domain (also known as the spectral domain), time scale, and time-frequency, amongst others. One of the main methods used for pattern recognition in myoelectric signals is the Support Vector Machines (SVM) technique whose primary function is to identify an n-dimensional hyperplane to separate a set of input feature points into different classes. This technique has the potential to recognize complex patterns and on several occasions, it has proven its worth when compared to other classifiers such as Artificial Neural Network (ANN), Linear Discriminant Analysis (LDA), and Principal Component Analysis(PCA). The key concepts underlying the SVM are (a) the hyperplane separator; (b) the kernel function; (c) the optimal separation hyperplane; and (d) a soft margin (hyperplane tolerance).

The paper utilizes the graph embeddings generated for entities of a large biomedical database to perform link prediction to capture various new relationships among different entities. A novel node similarity measure is proposed that utilizes the graph embeddings and link prediction scores to find similarity scores among various drugs which can be used by the medical experts to recommend alternative drugs to avoid side effects from original one. Utilizing machine learning on knowledge graph for drug similarity and recommendation will be less costly and less time consuming with higher scalability as compared to traditional biomedical methods due to the dependency on costly medical equipment and experts of the latter ones.

Vision transformers (ViTs) have pushed the state-of-the-art for various visual recognition tasks by patch-wise image tokenization followed by self-attention. However, the employment of self-attention modules results in a quadratic complexity in both computation and memory usage. Various attempts on approximating the self-attention computation with linear complexity have been made in Natural Language Processing. However, an in-depth analysis in this work shows that they are either theoretically flawed or empirically ineffective for visual recognition. We further identify that their limitations are rooted in keeping the softmax self-attention during approximations. Specifically, conventional self-attention is computed by normalizing the scaled dot-product between token feature vectors. Keeping this softmax operation challenges any subsequent linearization efforts. Based on this insight, for the first time, a softmax-free transformer or SOFT is proposed. To remove softmax in self-attention, Gaussian kernel function is used to replace the dot-product similarity without further normalization. This enables a full self-attention matrix to be approximated via a low-rank matrix decomposition. The robustness of the approximation is achieved by calculating its Moore-Penrose inverse using a Newton-Raphson method. Extensive experiments on ImageNet show that our SOFT significantly improves the computational efficiency of existing ViT variants. Crucially, with a linear complexity, much longer token sequences are permitted in SOFT, resulting in superior trade-off between accuracy and complexity.

With the explosive growth of information technology, multi-view graph data have become increasingly prevalent and valuable. Most existing multi-view clustering techniques either focus on the scenario of multiple graphs or multi-view attributes. In this paper, we propose a generic framework to cluster multi-view attributed graph data. Specifically, inspired by the success of contrastive learning, we propose multi-view contrastive graph clustering (MCGC) method to learn a consensus graph since the original graph could be noisy or incomplete and is not directly applicable. Our method composes of two key steps: we first filter out the undesirable high-frequency noise while preserving the graph geometric features via graph filtering and obtain a smooth representation of nodes; we then learn a consensus graph regularized by graph contrastive loss. Results on several benchmark datasets show the superiority of our method with respect to state-of-the-art approaches. In particular, our simple approach outperforms existing deep learning-based methods.

With the widespread use of machine learning to support decision-making, it is increasingly important to verify and understand the reasons why a particular output is produced. Although post-training feature importance approaches assist this interpretation, there is an overall lack of consensus regarding how feature importance should be quantified, making explanations of model predictions unreliable. In addition, many of these explanations depend on the specific machine learning approach employed and on the subset of data used when calculating feature importance. A possible solution to improve the reliability of explanations is to combine results from multiple feature importance quantifiers from different machine learning approaches coupled with re-sampling. Current state-of-the-art ensemble feature importance fusion uses crisp techniques to fuse results from different approaches. There is, however, significant loss of information as these approaches are not context-aware and reduce several quantifiers to a single crisp output. More importantly, their representation of 'importance' as coefficients is misleading and incomprehensible to end-users and decision makers. Here we show how the use of fuzzy data fusion methods can overcome some of the important limitations of crisp fusion methods.

Assessing the personality of software engineers may help to match individual traits with the characteristics of development activities such as code review and testing, as well as support managers in team composition. However, self-assessment questionnaires are not a practical solution for collecting multiple observations on a large scale. Instead, automatic personality detection, while overcoming these limitations, is based on off-the-shelf solutions trained on non-technical corpora, which might not be readily applicable to technical domains like Software Engineering (SE). In this paper, we first assess the performance of general-purpose personality detection tools when applied to a technical corpus of developers' emails retrieved from the public archives of the Apache Software Foundation. We observe a general low accuracy of predictions and an overall disagreement among the tools. Second, we replicate two previous research studies in SE by replacing the personality detection tool used to infer developers' personalities from pull-request discussions and emails. We observe that the original results are not confirmed, i.e., changing the tool used in the original study leads to diverging conclusions. Our results suggest a need for personality detection tools specially targeted for the software engineering domain.

Many functions have approximately-known upper and/or lower bounds, potentially aiding the modeling of such functions. In this paper, we introduce Gaussian process models for functions where such bounds are (approximately) known. More specifically, we propose the first use of such bounds to improve Gaussian process (GP) posterior sampling and Bayesian optimization (BO). That is, we transform a GP model satisfying the given bounds, and then sample and weight functions from its posterior. To further exploit these bounds in BO settings, we present bounded entropy search (BES) to select the point gaining the most information about the underlying function, estimated by the GP samples, while satisfying the output constraints. We characterize the sample variance bounds and show that the decision made by BES is explainable. Our proposed approach is conceptually straightforward and can be used as a plug in extension to existing methods for GP posterior sampling and Bayesian optimization.