The Cox Proportional Hazards (CPH) model has long been the preferred survival model for its explainability. However, to increase its predictive power beyond its linear log-risk, it was extended to utilize deep neural networks, sacrificing its explainability. In this work, we explore the potential of self-explaining neural networks (SENN) for survival analysis. We propose a new locally explainable Cox proportional hazards model, named CoxSE, by estimating a locally-linear log-hazard function using the SENN. We also propose a modification to the Neural additive (NAM) model, hybrid with SENN, named CoxSENAM, which enables the control of the stability and consistency of the generated explanations. Several experiments using synthetic and real datasets are presented, benchmarking CoxSE and CoxSENAM against a NAM-based model, a DeepSurv model explained with SHAP, and a linear CPH model. The results show that, unlike the NAM-based model, the SENN-based model can provide more stable and consistent explanations while maintaining the predictive power of the black-box model. The results also show that, due to their structural design, NAM-based models demonstrate better robustness to non-informative features. Among the models, the hybrid model exhibits the best robustness. Full implementation is available on GitHub. © 2025 The Authors

Activity recognition of industrial equipment is an enabler for IoT-based monitoring and optimization of material handling operations. This paper investigates forklift activity recognition using onboard Controller Area Network (CAN) data and state-of-the-art time series classification models, with a focus on generalization across industrial scenarios. Data were collected from forklifts operating in different warehouses under varying operational conditions, and several models were evaluated under within-scenario and cross-scenario settings. The results show that good recognition performance is achieved when training and testing within the same scenario, confirming that onboard CAN data provide a sufficient sensing modality for activity classification. However, performance degrades significantly when models are applied to unseen warehouses and operators, revealing a critical generalization gap that limits scalability of supervised approaches. These findings highlight an important deployment challenge and demonstrate the need for methods that support robust cross-site adaptation. The results provide practical insight into the feasibility and limitations of CAN-based activity recognition and establish a foundation for developing lowcost, scalable, and deployable monitoring solutions for industrial material handling applications.

Synthetic Electronic Health Records (EHRs) provide a viable means of accessing EHR data while addressing the privacy concerns related to the use of EHRs. A key characteristic of EHRs is the irregular timing of clinical events, admissions, and associated temporal trends. Many existing models for generating synthetic EHRs overlook these temporal irregularities, often assuming uniform intervals between clinical events for each patient and neglecting the time component, which hinders the representation of true temporal dynamics. To address these limitations, we propose TempEHR, a framework designed to synthesise EHRs, emphasising temporal awareness. We employ a time-aware Variational Autoencoder (VAE), specifically a Maximum Mean Discrepancy VAE (MMD-VAE), leveraging Time-aware Long Short-Term Memory (T-LSTM) layers to generate temporal synthetic EHRs along with time information. Simultaneously, we enhance the temporal awareness of our proposed model with a novel network we refer to as a TrendFinder. TrendFinder leverages a moving average to extract the temporal patterns inherent in irregular longitudinal EHR data. This approach seeks to enhance the fidelity and usefulness of synthetic EHRs for research and clinical applications. We assess the effectiveness of TempEHR using EHRs from the Medical Information Mart for Intensive Care (MIMIC-IV) repository. Our results demonstrate the potential of the proposed method in capturing the temporal patterns present in EHRs in utility, fidelity and privacy evaluations. © 2026 The Author(s), under exclusive license to Springer Nature Switzerland AG

A comprehensive review of Machine Activity Recognition (MAR) is presented, a field dedicated to recognizing the operational states of industrial machines through sensor data and artificial intelligence (AI) techniques. Covering 133 technical papers published in the period 2004 to 2025, the review categorizes and summarizes MAR research by equipment type, sensor modality, and algorithmic approach. It charts the progression from rule-based and classical machine learning methods to more sophisticated deep learning and self-supervised techniques. Since 2019, the field has shifted towards deep learning, driven by greater data availability and the need for models capable of addressing complex applications. The use of internal machine signals has also notably increased during this period. In addition to organizing a timeline of MAR development, the paper discusses machine learning challenges within the MAR context, such as inconsistent activity definitions, generalization across scenarios, imbalanced data, and the underutilization of unlabeled data in real-world applications. It further emphasizes the importance of aligning algorithm development with practical deployment requirements. By synthesizing research developments over time and highlighting emerging trends, this work aims to support the advancement of AI for MAR toward robust, scalable, and real-world solutions. © 2026 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/

In the field of Human Activity Recognition (HAR), the rapid evolution of wearable devices necessitates models that are generalizable and can adapt to entirely new subjects and activities with very limited labeled data. Conventional deep learning models, constrained by their reliance on large training datasets and limited adaptability to novel scenarios, face challenges in these settings. This paper introduces a novel few-shot HAR strategy employing meta-learning, which facilitates rapid adaptation to unseen subjects and activities using minimal annotated samples. Our method augments time series data with a range of transformations, each assigned a learnable weight, enabling the model to prioritize the most effective augmentations and discard the irrelevant ones. Throughout the meta-training phase, the model learns to identify an optimally weighted combination of these transformations, significantly improving the model’s adaptability and generalization to new situations with scarce labeled data. During meta-testing, this knowledge enables the model to efficiently learn from and adapt to a very limited set of labeled samples from completely new subjects undertaking entirely new activities. Extensive experiments on various HAR datasets demonstrate our method’s enhanced adaptability and generalization to tasks never encountered during training, achieving a performance improvement of up to  across all datasets. These results affirm MADA’s potential for real-world applications characterized by limited data availability. The Author(s) 2025

[No abstract available]

Meta-learning empowers learning systems with the ability to acquire knowledge from multiple tasks, enabling faster adaptation and generalization to new tasks. This review provides a comprehensive technical overview of meta-learning, emphasizing its importance in real-world applications where data may be scarce or expensive to obtain. The paper covers the state-of-the-art meta-learning approaches and explores the relationship between meta-learning and multi-task learning, transfer learning, domain adaptation and generalization, selfsupervised learning, personalized federated learning, and continual learning. By highlighting the synergies between these topics and the field of meta-learning, the paper demonstrates how advancements in one area can benefit the field as a whole, while avoiding unnecessary duplication of efforts. Additionally, the paper delves into advanced meta-learning topics such as learning from complex multi-modal task distributions, unsupervised metalearning, learning to efficiently adapt to data distribution shifts, and continual meta-learning. Lastly, the paper highlights open problems and challenges for future research in the field. By synthesizing the latest research developments, this paper provides a thorough understanding of meta-learning and its potential impact on various machine learning applications. We believe that this technical overview will contribute to the advancement of meta-learning and its practical implications in addressing realworld problems.

Synthetic data generation has been proposed as a potential solution to accessing Electronic Health Records (EHRs) while minimizing the privacy risks associated with real EHRs. Nevertheless, the practical use of synthetic EHRs rests on their ability to resemble the quality of real EHRs. Existing evaluations of synthetic EHRs often focus on assessing them as static snapshots frozen in time, neglecting temporal dependencies and varying temporal patterns. Moreover, some of these methods rely on subjective judgments, are limited to segmentable time-series, and employ methods that adopt a one-to-one approach. This study employs a comprehensive approach to evaluating fidelity in synthetic time-series EHRs to address these challenges. We extend the functionality of time-series analysis methods such as temporal clustering, time-series similarity measures, Sample Entropy, and trend analysis, to evaluate varying temporal patterns in synthetic time-series EHRs. Our findings provide valuable insights into how synthetic EHRs align with real EHRs in the temporal context, considering aspects such as patient groupings, temporal dynamics, predictability, and directional change. We empirically demonstrate the feasibility of assessing temporal fidelity with these methods, offering an understanding of the quality of synthetic EHRs in capturing the varying temporal patterns inherent in EHRs. © 2024 IEEE.

Recent studies have shown how synthetic data generation methods can be applied to electronic health records (EHRs) to obtain synthetic versions that do not violate privacy rules. This growing body of research has resulted in the emergence of numerous methods for evaluating the quality of generated data, with new publications often introducing novel evaluation methods. This work presents a detailed review of synthetic EHRs, focusing on the various evaluation methods used to assess the quality of the generated EHRs. We discuss the existing evaluation methods, offering insights into their use as well as providing an interpretation of the evaluation metrics from the perspectives of achieving fidelity, utility and privacy. Furthermore, we highlight the key factors influencing the selection of evaluation methods, such as the type of data (e.g., categorical, continuous, or discrete) and the mode of application (e.g., patient level, cohort level, and feature level). To assess the effectiveness of current evaluation measures, we conduct a series of experiments to shed light on the potential limitations of these measures. The findings from these experiments reveal notable shortcomings, including the need for meticulous application of methods to the data to reduce inconsistent evaluations, the qualitative nature of some assessments subject to individual judgment, the need for clinical validations, and the absence of techniques to evaluate temporal dependencies within the data. This highlights the need to place greater emphasis on evaluation measures, their application, and the development of comprehensive evaluation frameworks as it is crucial for advancing progress in this field. © 2024 The Author(s)