This paper proposes a novel graph-based framework for robust and interpretable multiclass fault diagnosis in rotating machinery. The method integrates entropy-optimized signal segmentation, time-frequency feature extraction, and graph-theoretic modeling to transform vibration signals into structured representations suitable for classification. Graph metrics, such as average shortest path length, modularity, and spectral gap, are computed and combined with local features to capture global and segment-level fault characteristics. The proposed method achieves high diagnostic accuracy when evaluated on two benchmark datasets, the CWRU bearing dataset (under 0-3 HP loads) and the SU gearbox and bearing datasets (under different speed-load configurations). Classification scores reach up to 99.8% accuracy on Case Western Reserve University (CWRU) and 100% accuracy on the Southeast University datasets using a logistic regression classifier. Furthermore, the model exhibits strong noise resilience, maintaining over 95.4% accuracy at high noise levels (standard deviation = 0.5), and demonstrates excellent cross-domain transferability with up to 99.7% F1-score in load-transfer scenarios. Compared to traditional techniques, this approach requires no deep learning architecture, enabling lower complexity while ensuring interpretability. The results confirm the method's scalability, reliability, and potential for real-time deployment in industrial diagnostics.

Personal use of this material is permitted. This paper has been published in the 8th IEEE Conference on Industrial Cyber-Physical Systems (ICPS) in Emden, Germany, May 12-15, 2025. Abstract --Accurately predicting the lifespan of critical device components is essential for maintenance planning and production optimization, making it a topic of significant interest in both academia and industry. In this work, we investigate the use of survival analysis for predicting the lifespan of production printheads developed by Canon Production Printing. Specifically, we focus on the application of five techniques to estimate survival probabilities and failure rates: the Kaplan-Meier estimator, Cox proportional hazard model, Weibull accelerated failure time model, random survival forest, and gradient boosting. The resulting estimates are further refined using isotonic regression and subsequently aggregated to determine the expected number of failures. The predictions are then validated against real-world ground truth data across multiple time windows to assess model reliability. Our quantitative evaluation using three performance metrics demonstrates that survival analysis outperforms industry-standard baseline methods for printhead lifespan prediction.

This paper studies the problem of dimension reduction, tailored to improving time series forecasting with high-dimensional predictors. We propose a novel Supervised Deep Dynamic Principal component analysis (SDDP) framework that incorporates the target variable and lagged observations into the factor extraction process. Assisted by a temporal neural network, we construct target-aware predictors by scaling the original predictors in a supervised manner, with larger weights assigned to predictors with stronger forecasting power. A principal component analysis is then performed on the target-aware predictors to extract the estimated SDDP factors. This supervised factor extraction not only improves predictive accuracy in the downstream forecasting task but also yields more interpretable and target-specific latent factors. Building upon SDDP, we propose a factor-augmented nonlinear dynamic forecasting model that unifies a broad family of factor-model-based forecasting approaches. To further demonstrate the broader applicability of SDDP, we extend our studies to a more challenging scenario when the predictors are only partially observable. We validate the empirical performance of the proposed method on several real-world public datasets. The results show that our algorithm achieves notable improvements in forecasting accuracy compared to state-of-the-art methods.

Negative binomial regression is essential for analyzing over-dispersed count data in in comparative studies, but parameter estimation becomes computationally challenging in large screens requiring millions of comparisons. We investigate using a pre-trained transformer to produce estimates of negative binomial regression parameters from observed count data, trained through synthetic data generation to learn to invert the process of generating counts from parameters. The transformer method achieved better parameter accuracy than maximum likelihood optimization while being 20 times faster. However, comparisons unexpectedly revealed that method of moment estimates performed as well as maximum likelihood optimization in accuracy, while being 1,000 times faster and producing better-calibrated and more powerful tests, making it the most efficient solution for this application.

Distance-based classifiers, such as k-nearest neighbors and support vector machines, continue to be a workhorse of machine learning, widely used in science and industry. In practice, to derive insights from these models, it is also important to ensure that their predictions are explainable. While the field of Explainable AI has supplied methods that are in principle applicable to any model, it has also emphasized the usefulness of latent structures (e.g. the sequence of layers in a neural network) to produce explanations. In this paper, we contribute by uncovering a hidden neural network structure in distance-based classifiers (consisting of linear detection units combined with nonlinear pooling layers) upon which Explainable AI techniques such as layer-wise relevance propagation (LRP) become applicable. Through quantitative evaluations, we demonstrate the advantage of our novel explanation approach over several baselines. We also show the overall usefulness of explaining distance-based models through two practical use cases.

Active Feature Acquisition is an instance-wise, sequential decision making problem. The aim is to dynamically select which feature to measure based on current observations, independently for each test instance. Common approaches either use Reinforcement Learning, which experiences training difficulties, or greedily maximize the conditional mutual information of the label and unobserved features, which makes myopic acquisitions. To address these shortcomings, we introduce a latent variable model, trained in a supervised manner. Acquisitions are made by reasoning about the features across many possible unobserved realizations in a stochastic latent space. Extensive evaluation on a large range of synthetic and real datasets demonstrates that our approach reliably outperforms a diverse set of baselines.

Trustworthy interpretation of deep learning models is critical for neuroimaging applications, yet commonly used Explainable AI (XAI) methods lack rigorous validation, risking misinterpretation. We performed the first large-scale, systematic comparison of XAI methods on ~45,000 structural brain MRIs using a novel XAI validation framework. This framework establishes verifiable ground truth by constructing prediction tasks with known signal sources - from localized anatomical features to subject-specific clinical lesions - without artificially altering input images. Our analysis reveals systematic failures in two of the most widely used methods: GradCAM consistently failed to localize predictive features, while Layer-wise Relevance Propagation generated extensive, artifactual explanations that suggest incompatibility with neuroimaging data characteristics. Our results indicate that these failures stem from a domain mismatch, where methods with design principles tailored to natural images require substantial adaptation for neuroimaging data. In contrast, the simpler, gradient-based method SmoothGrad, which makes fewer assumptions about data structure, proved consistently accurate, suggesting its conceptual simplicity makes it more robust to this domain shift. These findings highlight the need for domain-specific adaptation and validation of XAI methods, suggest that interpretations from prior neuroimaging studies using standard XAI methodology warrant re-evaluation, and provide urgent guidance for practical application of XAI in neuroimaging.

Abstract--The accurate labeling of datasets is often both costly and time-consuming. Given an unlabeled dataset, programma tic weak supervision obtains probabilistic predictions for th e labels by leveraging multiple weak labeling functions (LFs) that p ro-vide rough guesses for labels. Weak LFs commonly provide guesses with assorted types and unknown interdependences that can result in unreliable predictions. This paper presents a methodology for programma tic weak supervision that can provide confidence intervals for l abel probabilities and obtain more reliable predictions. In par ticular, the methods proposed use uncertainty sets of distributions that encapsulate the information provided by LFs with unrestric ted behavior and typology. Experiments on multiple benchmark datasets show the improvement of the presented methods over the state-of-the-art and the practicality of the confidence intervals presented. OR many machine learning applications, the accurate labeling of datasets is both costly and time-consuming [1]-[4]. Given an unlabeled dataset, methods for programmatic weak supervision aim to leverage multiple wea k labeling functions (LFs) to provide accurate labels [5], [6 ]. Since common LFs only provide rough guesses for labels, programmatic weak supervision methods use the outputs of multiple LFs to obtain probabilistic predictions for the la bel of each instance [7]-[13]. These predictions can then be use d to create a fully supervised dataset composed by the instanc es corresponding to high-confidence predictions, e.g., a labe l with a large enough predicted probability is regarded as the actu al Manuscript received September 30, 2024; accepted August 4, 2025.

Existing work has shown that cross-validation (CV) can be used to provide an asymptotic confidence interval for the test error of a stable machine learning algorithm, and existing stability results for many popular algorithms can be applied to derive positive instances where such confidence intervals will be valid. However, in the common setting where CV is used to compare two algorithms, it becomes necessary to consider a notion of relative stability which cannot easily be derived from existing stability results, even for simple algorithms. To better understand relative stability and when CV provides valid confidence intervals for the test error difference of two algorithms, we study the soft-thresholded least squares algorithm, a close cousin of the Lasso. We prove that while stability holds when assessing the individual test error of this algorithm, relative stability fails to hold when comparing the test error of two such algorithms, even in a sparse low-dimensional linear model setting. Additionally, we empirically confirm the invalidity of CV confidence intervals for the test error difference when either soft-thresholding or the Lasso is used. In short, caution is needed when quantifying the uncertainty of CV estimates of the performance difference of two machine learning algorithms, even when both algorithms are individually stable.

Latent structure methods, specifically linear continuous latent structure methods, are a type of fundamental statistical learning strategy. They are widely used for dimension reduction, regression and prediction, in the fields of chemometrics, economics, social science and etc. However, due to the lack of model inference, generative form, and unidentifiable parameters, most of these methods are always used as an algorithm, instead of a model. This paper proposed a Generative Flexible Latent Structure Regression (GFLSR) model structure to address this problem. Moreover, we show that most linear continuous latent variable methods can be represented under the proposed framework. The recursive structure allows potential model inference and residual analysis. Then, the traditional Partial Least Squares (PLS) is focused; we show that the PLS can be specialised in the proposed model structure, named Generative-PLS. With a model structure, we analyse the convergence of the parameters and the latent variables. Under additional distribution assumptions, we show that the proposed model structure can lead to model inference without solving the probabilistic model. Additionally, we proposed a novel bootstrap algorithm that enables uncertainty on parameters and on prediction for new datasets. A simulation study and a Real-world dataset are used to verify the proposed Generative-PLS model structure. Although the traditional PLS is a special case, this proposed GFLSRM structure leads to a potential inference structure for all the linear continuous latent variable methods.