Data augmentation has been pivotal in successfully training deep learning models on classification tasks over the past decade. An important subclass of data augmentation techniques - which includes both label smoothing and Mixup - involves modifying not only the input data but also the input label during model training. In this work, we analyze the role played by the label augmentation aspect of such methods. We first prove that linear models on binary classification data trained with label augmentation learn only the minimum variance features in the data, while standard training (which includes weight decay) can learn higher variance features. We then use our techniques to show that even for nonlinear models and general data distributions, the label smoothing and Mixup losses are lower bounded by a function of the model output variance. An important consequence of our results is negative: label smoothing and Mixup can be less robust to spurious correlations in the data. We verify that our theory reflects practice via experiments on image classification benchmarks modified to have spurious correlations.
For the Facial Action Unit (AU) detection task, accurately capturing the subtle facial differences between distinct AUs is essential for reliable detection. Additionally, AU detection faces challenges from class imbalance and the presence of noisy or false labels, which undermine detection accuracy. In this paper, we introduce a novel contrastive learning framework aimed for AU detection that incorporates both self-supervised and supervised signals, thereby enhancing the learning of discriminative features for accurate AU detection. To tackle the class imbalance issue, we employ a negative sample re-weighting strategy that adjusts the step size of updating parameters for minority and majority class samples. Moreover, to address the challenges posed by noisy and false AU labels, we employ a sampling technique that encompasses three distinct types of positive sample pairs. This enables us to inject self-supervised signals into the supervised signal, effectively mitigating the adverse effects of noisy labels. Our experimental assessments, conducted on five widely-utilized benchmark datasets (BP4D, DISFA, BP4D+, GFT and Aff-Wild2), underscore the superior performance of our approach compared to state-of-the-art methods of AU detection.
In this paper, a new system based on combinations of a shape descriptor and a contour descriptor has been proposed for classifying inserts in milling processes according to their wear level following a computer vision based approach. To describe the wear region shape we have proposed a new descriptor called ShapeFeat and its contour has been characterized using the method BORCHIZ that, to the best of our knowledge, achieves the best performance for tool wear monitoring following a computer vision-based approach. Results show that the combination of BORCHIZ with ShapeFeat using a late fusion method improves the classification performance significantly, obtaining an accuracy of 91.44% in the binary classification (i.e. the classification of the wear as high or low) and 82.90% using three target classes (i.e. classification of the wear as high, medium or low). These results outperform the ones obtained by both descriptors used on their own, which achieve accuracies of 88.70 and 80.67% for two and three classes, respectively, using ShapeFeat and 87.06 and 80.24% with B-ORCHIZ. This study yielded encouraging results for the manufacturing community in order to classify automatically the inserts in terms of their wear for milling processes.
We study the fundamental optimization principles of self-attention, the defining mechanism of transformers, by analyzing the implicit bias of gradient-based optimizers in training a self-attention layer with a linear decoder in binary classification. Building on prior studies in linear logistic regression, recent findings demonstrate that the key-query matrix $W_t$ from gradient-descent (GD) converges in direction towards $W_{mm}$, which maximizes the margin between optimal and non-optimal tokens across sequences. However, this convergence is local, dependent on initial conditions, only holds asymptotically as the number of iterations increases, and leaves questions about the potential benefits of adaptive step-size rules unaddressed. To bridge this gap, we first establish scenarios for which convergence is provably \emph{global}. We then analyze two adaptive step-size strategies: normalized GD and Polyak step-size, demonstrating \emph{finite-time} convergence rates for $W_t$ to $W_{mm}$, and quantifying the sparsification rate of the attention map. These findings not only show that these strategies can accelerate parameter convergence over standard GD in a non-convex setting but also deepen the understanding of the implicit bias in self-attention, linking it more closely to the phenomena observed in linear logistic regression despite its intricate non-convex nature.
Although binary classification is a well-studied problem in computer vision, training reliable classifiers under severe class imbalance remains a challenging problem. Recent work has proposed techniques that mitigate the effects of training under imbalance by modifying the loss functions or optimization methods. While this work has led to significant improvements in the overall accuracy in the multi-class case, we observe that slight changes in hyperparameter values of these methods can result in highly variable performance in terms of Receiver Operating Characteristic (ROC) curves on binary problems with severe imbalance. To reduce the sensitivity to hyperparameter choices and train more general models, we propose training over a family of loss functions, instead of a single loss function. We develop a method for applying Loss Conditional Training (LCT) to an imbalanced classification problem. Extensive experiment results, on both CIFAR and Kaggle competition datasets, show that our method improves model performance and is more robust to hyperparameter choices. Code is available at https://github.com/klieberman/roc_lct.
We have widely observed that neural networks are vulnerable to small additive perturbations to the input causing misclassification. In this paper, we focus on the $\ell_0$-bounded adversarial attacks, and aim to theoretically characterize the performance of adversarial training for an important class of truncated classifiers. Such classifiers are shown to have strong performance empirically, as well as theoretically in the Gaussian mixture model, in the $\ell_0$-adversarial setting. The main contribution of this paper is to prove a novel generalization bound for the binary classification setting with $\ell_0$-bounded adversarial perturbation that is distribution-independent. Deriving a generalization bound in this setting has two main challenges: (i) the truncated inner product which is highly non-linear; and (ii) maximization over the $\ell_0$ ball due to adversarial training is non-convex and highly non-smooth. To tackle these challenges, we develop new coding techniques for bounding the combinatorial dimension of the truncated hypothesis class.
Power grids are critical infrastructures of paramount importance to modern society and, therefore, engineered to operate under diverse conditions and failures. The ongoing energy transition poses new challenges for the decision-makers and system operators. Therefore, developing grid analysis algorithms is important for supporting reliable operations. These key tools include power flow analysis and system security analysis, both needed for effective operational and strategic planning. The literature review shows a growing trend of machine learning (ML) models that perform these analyses effectively. In particular, Graph Neural Networks (GNNs) stand out in such applications because of the graph-based structure of power grids. However, there is a lack of publicly available graph datasets for training and benchmarking ML models in electrical power grid applications. First, we present PowerGraph, which comprises GNN-tailored datasets for i) power flows, ii) optimal power flows, and iii) cascading failure analyses of power grids. Second, we provide ground-truth explanations for the cascading failure analysis. Finally, we perform a complete benchmarking of GNN methods for node-level and graph-level tasks and explainability. Overall, PowerGraph is a multifaceted GNN dataset for diverse tasks that includes power flow and fault scenarios with real-world explanations, providing a valuable resource for developing improved GNN models for node-level, graph-level tasks and explainability methods in power system modeling. The dataset is available at https://figshare.com/articles/dataset/PowerGraph/22820534 and the code at https://github.com/PowerGraph-Datasets.
The bispectrum stands out as a revolutionary tool in frequency domain analysis, leaping the usual power spectrum by capturing crucial phase information between frequency components. In our innovative study, we have utilized the bispectrum to analyze and decode complex grasping movements, gathering EEG data from five human subjects. We put this data through its paces with three classifiers, focusing on both magnitude and phase-related features. The results highlight the bispectrum's incredible ability to delve into neural activity and differentiate between various grasping motions with the Support Vector Machine (SVM) classifier emerging as a standout performer. In binary classification, it achieved a remarkable 97\% accuracy in identifying power grasp, and in the more complex multiclass tasks, it maintained an impressive 94.93\% accuracy. This finding not only underscores the bispectrum's analytical strength but also showcases the SVM's exceptional capability in classification, opening new doors in our understanding of movement and neural dynamics.
Despite the predominance of contextualized embeddings in NLP, approaches to detect semantic change relying on these embeddings and clustering methods underperform simpler counterparts based on static word embeddings. This stems from the poor quality of the clustering methods to produce sense clusters -- which struggle to capture word senses, especially those with low frequency. This issue hinders the next step in examining how changes in word senses in one language influence another. To address this issue, we propose a graph-based clustering approach to capture nuanced changes in both high- and low-frequency word senses across time and languages, including the acquisition and loss of these senses over time. Our experimental results show that our approach substantially surpasses previous approaches in the SemEval2020 binary classification task across four languages. Moreover, we showcase the ability of our approach as a versatile visualization tool to detect semantic changes in both intra-language and inter-language setups. We make our code and data publicly available.
Invasive Coronary Angiography (ICA) images are considered the gold standard for assessing the state of the coronary arteries. Deep learning classification methods are widely used and well-developed in different areas where medical imaging evaluation has an essential impact due to the development of computer-aided diagnosis systems that can support physicians in their clinical procedures. In this paper, a new performance analysis of deep learning methods for binary ICA classification with different lesion degrees is reported. To reach this goal, an annotated dataset of ICA images that contains the ground truth, the location of lesions and seven possible severity degrees ranging between 0% and 100% was employed. The ICA images were divided into 'lesion' or 'non-lesion' patches. We aim to study how binary classification performance is affected by the different lesion degrees considered in the positive class. Therefore, five known convolutional neural network architectures were trained with different input images where different lesion degree ranges were gradually incorporated until considering the seven lesion degrees. Besides, four types of experiments with and without data augmentation were designed, whose F-measure and Area Under Curve (AUC) were computed. Reported results achieved an F-measure and AUC of 92.7% and 98.1%, respectively. However, lesion classification is highly affected by the degree of the lesion intended to classify, with 15% less accuracy when <99% lesion patches are present.
State-of-the-art deepfake detection approaches rely on image-based features extracted via neural networks. While these approaches trained in a supervised manner extract likely fake features, they may fall short in representing unnatural `non-physical' semantic facial attributes -- blurry hairlines, double eyebrows, rigid eye pupils, or unnatural skin shading. However, such facial attributes are easily perceived by humans and used to discern the authenticity of an image based on human common sense. Furthermore, image-based feature extraction methods that provide visual explanations via saliency maps can be hard to interpret for humans. To address these challenges, we frame deepfake detection as a Deepfake Detection VQA (DD-VQA) task and model human intuition by providing textual explanations that describe common sense reasons for labeling an image as real or fake. We introduce a new annotated dataset and propose a Vision and Language Transformer-based framework for the DD-VQA task. We also incorporate text and image-aware feature alignment formulation to enhance multi-modal representation learning. As a result, we improve upon existing deepfake detection models by integrating our learned vision representations, which reason over common sense knowledge from the DD-VQA task. We provide extensive empirical results demonstrating that our method enhances detection performance, generalization ability, and language-based interpretability in the deepfake detection task.
The structure of data organization is widely recognized as having a substantial influence on the efficacy of machine learning algorithms, particularly in binary classification tasks. Our research provides a theoretical framework suggesting that the maximum potential of binary classifiers on a given dataset is primarily constrained by the inherent qualities of the data. Through both theoretical reasoning and empirical examination, we employed standard objective functions, evaluative metrics, and binary classifiers to arrive at two principal conclusions. Firstly, we show that the theoretical upper bound of binary classification performance on actual datasets can be theoretically attained. This upper boundary represents a calculable equilibrium between the learning loss and the metric of evaluation. Secondly, we have computed the precise upper bounds for three commonly used evaluation metrics, uncovering a fundamental uniformity with our overarching thesis: the upper bound is intricately linked to the dataset's characteristics, independent of the classifier in use. Additionally, our subsequent analysis uncovers a detailed relationship between the upper limit of performance and the level of class overlap within the binary classification data. This relationship is instrumental for pinpointing the most effective feature subsets for use in feature engineering.
We consider the optimization of complex performance metrics in multi-label classification under the population utility framework. We mainly focus on metrics linearly decomposable into a sum of binary classification utilities applied separately to each label with an additional requirement of exactly $k$ labels predicted for each instance. These "macro-at-$k$" metrics possess desired properties for extreme classification problems with long tail labels. Unfortunately, the at-$k$ constraint couples the otherwise independent binary classification tasks, leading to a much more challenging optimization problem than standard macro-averages. We provide a statistical framework to study this problem, prove the existence and the form of the optimal classifier, and propose a statistically consistent and practical learning algorithm based on the Frank-Wolfe method. Interestingly, our main results concern even more general metrics being non-linear functions of label-wise confusion matrices. Empirical results provide evidence for the competitive performance of the proposed approach.
The intricate relationship between human decision-making and emotions, particularly guilt and regret, has significant implications on behavior and well-being. Yet, these emotions subtle distinctions and interplay are often overlooked in computational models. This paper introduces a dataset tailored to dissect the relationship between guilt and regret and their unique textual markers, filling a notable gap in affective computing research. Our approach treats guilt and regret recognition as a binary classification task and employs three machine learning and six transformer-based deep learning techniques to benchmark the newly created dataset. The study further implements innovative reasoning methods like chain-of-thought and tree-of-thought to assess the models interpretive logic. The results indicate a clear performance edge for transformer-based models, achieving a 90.4% macro F1 score compared to the 85.3% scored by the best machine learning classifier, demonstrating their superior capability in distinguishing complex emotional states.
Training machine learning and statistical models often involves optimizing a data-driven risk criterion. The risk is usually computed with respect to the empirical data distribution, but this may result in poor and unstable out-of-sample performance due to distributional uncertainty. In the spirit of distributionally robust optimization, we propose a novel robust criterion by combining insights from Bayesian nonparametric (i.e., Dirichlet process) theory and a recent decision-theoretic model of smooth ambiguity-averse preferences. First, we highlight novel connections with standard regularized empirical risk minimization techniques, among which Ridge and LASSO regressions. Then, we theoretically demonstrate the existence of favorable finite-sample and asymptotic statistical guarantees on the performance of the robust optimization procedure. For practical implementation, we propose and study tractable approximations of the criterion based on well-known Dirichlet process representations. We also show that the smoothness of the criterion naturally leads to standard gradient-based numerical optimization. Finally, we provide insights into the workings of our method by applying it to a variety of tasks based on simulated and real datasets.
In the ambitious realm of space AI, the integration of federated learning (FL) with low Earth orbit (LEO) satellite constellations holds immense promise. However, many challenges persist in terms of feasibility, learning efficiency, and convergence. These hurdles stem from the bottleneck in communication, characterized by sporadic and irregular connectivity between LEO satellites and ground stations, coupled with the limited computation capability of satellite edge computing (SEC). This paper proposes a novel FL-SEC framework that empowers LEO satellites to execute large-scale machine learning (ML) tasks onboard efficiently. Its key components include i) personalized learning via divide-and-conquer, which identifies and eliminates redundant satellite images and converts complex multi-class classification problems to simple binary classification, enabling rapid and energy-efficient training of lightweight ML models suitable for IoT/edge devices on satellites; ii) orbital model retraining, which generates an aggregated "orbital model" per orbit and retrains it before sending to the ground station, significantly reducing the required communication rounds. We conducted experiments using Jetson Nano, an edge device closely mimicking the limited compute on LEO satellites, and a real satellite dataset. The results underscore the effectiveness of our approach, highlighting SEC's ability to run lightweight ML models on real and high-resolution satellite imagery. Our approach dramatically reduces FL convergence time by nearly 30 times, and satellite energy consumption down to as low as 1.38 watts, all while maintaining an exceptional accuracy of up to 96%.
The feature selection in a traditional binary classification algorithm is always used in the stage of dataset preprocessing, which makes the obtained features not necessarily the best ones for the classification algorithm, thus affecting the classification performance. For a traditional rule-based binary classification algorithm, classification rules are usually deterministic, which results in the fuzzy information contained in the rules being ignored. To do so, this paper employs iterative feature selection in fuzzy rule-based binary classification. The proposed algorithm combines feature selection based on fuzzy correlation family with rule mining based on biclustering. It first conducts biclustering on the dataset after feature selection. Then it conducts feature selection again for the biclusters according to the feedback of biclusters evaluation. In this way, an iterative feature selection framework is build. During the iteration process, it stops until the obtained bicluster meets the requirements. In addition, the rule membership function is introduced to extract vectorized fuzzy rules from the bicluster and construct weak classifiers. The weak classifiers with good classification performance are selected by Adaptive Boosting and the strong classifier is constructed by "weighted average". Finally, we perform the proposed algorithm on different datasets and compare it with other peers. Experimental results show that it achieves good classification performance and outperforms its peers.
In an era where biometric security serves as a keystone of modern identity verification systems, ensuring the authenticity of these biometric samples is paramount. Liveness detection, the capability to differentiate between genuine and spoofed biometric samples, stands at the forefront of this challenge. This research presents a comprehensive evaluation of liveness detection models, with a particular focus on their performance in cross-database scenarios, a test paradigm notorious for its complexity and real-world relevance. Our study commenced by meticulously assessing models on individual datasets, revealing the nuances in their performance metrics. Delving into metrics such as the Half Total Error Rate, False Acceptance Rate, and False Rejection Rate, we unearthed invaluable insights into the models' strengths and weaknesses. Crucially, our exploration of cross-database testing provided a unique perspective, highlighting the chasm between training on one dataset and deploying on another. Comparative analysis with extant methodologies, ranging from convolutional networks to more intricate strategies, enriched our understanding of the current landscape. The variance in performance, even among state-of-the-art models, underscored the inherent challenges in this domain. In essence, this paper serves as both a repository of findings and a clarion call for more nuanced, data-diverse, and adaptable approaches in biometric liveness detection. In the dynamic dance between authenticity and deception, our work offers a blueprint for navigating the evolving rhythms of biometric security.
Driven by an increasing prevalence of trackers, ever more IoT sensors, and the declining cost of computing power, geospatial information has come to play a pivotal role in contemporary predictive models. While enhancing prognostic performance, geospatial data also has the potential to perpetuate many historical socio-economic patterns, raising concerns about a resurgence of biases and exclusionary practices, with their disproportionate impacts on society. Addressing this, our paper emphasizes the crucial need to identify and rectify such biases and calibration errors in predictive models, particularly as algorithms become more intricate and less interpretable. The increasing granularity of geospatial information further introduces ethical concerns, as choosing different geographical scales may exacerbate disparities akin to redlining and exclusionary zoning. To address these issues, we propose a toolkit for identifying and mitigating biases arising from geospatial data. Extending classical fairness definitions, we incorporate an ordinal regression case with spatial attributes, deviating from the binary classification focus. This extension allows us to gauge disparities stemming from data aggregation levels and advocates for a less interfering correction approach. Illustrating our methodology using a Parisian real estate dataset, we showcase practical applications and scrutinize the implications of choosing geographical aggregation levels for fairness and calibration measures.
Classification is a fundamental task in many applications on which data-driven methods have shown outstanding performances. However, it is challenging to determine whether such methods have achieved the optimal performance. This is mainly because the best achievable performance is typically unknown and hence, effectively estimating it is of prime importance. In this paper, we consider binary classification problems and we propose an estimator for the false positive rate (FPR) of the Bayes classifier, that is, the optimal classifier with respect to accuracy, from a given dataset. Our method utilizes soft labels, or real-valued labels, which are gaining significant traction thanks to their properties. We thoroughly examine various theoretical properties of our estimator, including its consistency, unbiasedness, rate of convergence, and variance. To enhance the versatility of our estimator beyond soft labels, we also consider noisy labels, which encompass binary labels. For noisy labels, we develop effective FPR estimators by leveraging a denoising technique and the Nadaraya-Watson estimator. Due to the symmetry of the problem, our results can be readily applied to estimate the false negative rate of the Bayes classifier.