In multivariate time series (MTS) classification, finding the important features (e.g., sensors) for model performance is crucial yet challenging due to the complex, high-dimensional nature of MTS data, intricate temporal dynamics, and the necessity for domain-specific interpretations. Current explanation methods for MTS mostly focus on time-centric explanations, apt for pinpointing important time periods but less effective in identifying key features. This limitation underscores the pressing need for a feature-centric approach, a vital yet often overlooked perspective that complements time-centric analysis. To bridge this gap, our study introduces a novel feature-centric explanation and evaluation framework for MTS, named CAFO (Channel Attention and Feature Orthgonalization). CAFO employs a convolution-based approach with channel attention mechanisms, incorporating a depth-wise separable channel attention module (DepCA) and a QR decomposition-based loss for promoting feature-wise orthogonality. We demonstrate that this orthogonalization enhances the separability of attention distributions, thereby refining and stabilizing the ranking of feature importance. This improvement in feature-wise ranking enhances our understanding of feature explainability in MTS. Furthermore, we develop metrics to evaluate global and class-specific feature importance. Our framework's efficacy is validated through extensive empirical analyses on two major public benchmarks and real-world datasets, both synthetic and self-collected, specifically designed to highlight class-wise discriminative features. The results confirm CAFO's robustness and informative capacity in assessing feature importance in MTS classification tasks. This study not only advances the understanding of feature-centric explanations in MTS but also sets a foundation for future explorations in feature-centric explanations.

AI methods are finding an increasing number of applications, but their often black-box nature has raised concerns about accountability and trust. The field of explainable artificial intelligence (XAI) has emerged in response to the need for human understanding of AI models. Evolutionary computation (EC), as a family of powerful optimization and learning tools, has significant potential to contribute to XAI. In this paper, we provide an introduction to XAI and review various techniques in current use for explaining machine learning (ML) models. We then focus on how EC can be used in XAI, and review some XAI approaches which incorporate EC techniques. Additionally, we discuss the application of XAI principles within EC itself, examining how these principles can shed some light on the behavior and outcomes of EC algorithms in general, on the (automatic) configuration of these algorithms, and on the underlying problem landscapes that these algorithms optimize. Finally, we discuss some open challenges in XAI and opportunities for future research in this field using EC. Our aim is to demonstrate that EC is well-suited for addressing current problems in explainability and to encourage further exploration of these methods to contribute to the development of more transparent and trustworthy ML models and EC algorithms.

Accurately predicting heart activity and other biological signals is crucial for diagnosis and monitoring. Given that speech is an outcome of multiple physiological systems, a significant body of work studied the acoustic correlates of heart activity. Recently, self-supervised models have excelled in speech-related tasks compared to traditional acoustic methods. However, the robustness of data-driven representations in predicting heart activity remained unexplored. In this study, we demonstrate that self-supervised speech models outperform acoustic features in predicting heart activity parameters. We also emphasize the impact of individual variability on model generalizability. These findings underscore the value of data-driven representations in such tasks and the need for more speech-based physiological data to mitigate speaker-related challenges.

With the integration of advanced technologies and automation systems in manufacturing, available data formats are evolving to include complex data streams, such as sequences of images and videos, which can provide valuable information on the state of the machines and their components. The timely identification and diagnosis of faults can prevent equipment failure, reduce maintenance costs, improve system performance, and enhance safety. In recent years, Machine Learning (ML) algorithms, including Deep Learning (DL) models, have shown great promise in automating fault detection and diagnosis tasks. However, these models are often viewed as black boxes, making it challenging to understand how they arrived at their predictions. This lack of transparency can pose a significant barrier to adopting machine learning in safety-critical applications, where the interpretability and trustworthiness of the model are essential.

As a leading cause of mortality among women globally, the precise and timely diagnosis of breast cancer remains imperative for optimizing patient outcomes. While traditional diagnostic methodologies [2] have historically relied heavily on uni-modal approaches, the evolving landscape of medical data analytics underscores the significance of integrating diverse data sources beyond conventional imaging modalities [3]. Figure 1 illustrates a generic model for breast cancer diagnosis within the Computer-Aided Detection (CAD) framework. As depicted in Figure 2, breast cancer detection can be performed using various data types, employing either unimodal or multimodal approaches. The process initiates with data pre-processing, followed by feature extraction. To enhance the learning of feature representations from image data, segmentation may be conducted prior to feature extraction. Subsequently, the detection model is applied to generate a diagnosis from the processed data. Based on this diagnosis, further analyses are performed, including sub-type classification, grade classification, recurrence and metastasis prediction, as well as the incorporation of crowdsourcing and human-in-the-loop methodologies. These steps culminate in a final decision that informs subsequent treatment and monitoring strategies.

Integrating machine learning into Automated Control Systems (ACS) enhances decision-making in industrial process management. One of the limitations to the widespread adoption of these technologies in industry is the vulnerability of neural networks to adversarial attacks. This study explores the threats in deploying deep learning models for fault diagnosis in ACS using the Tennessee Eastman Process dataset. By evaluating three neural networks with different architectures, we subject them to six types of adversarial attacks and explore five different defense methods. Our results highlight the strong vulnerability of models to adversarial samples and the varying effectiveness of defense strategies. We also propose a novel protection approach by combining multiple defense methods and demonstrate it's efficacy. This research contributes several insights into securing machine learning within ACS, ensuring robust fault diagnosis in industrial processes.

Fault detection is crucial in industrial systems to prevent failures and optimize performance by distinguishing abnormal from normal operating conditions. Data-driven methods have been gaining popularity for fault detection tasks as the amount of condition monitoring data from complex industrial systems increases. Despite these advances, early fault detection remains a challenge under real-world scenarios. The high variability of operating conditions and environments makes it difficult to collect comprehensive training datasets that can represent all possible operating conditions, especially in the early stages of system operation. Furthermore, these variations often evolve over time, potentially leading to entirely new data distributions in the future that were previously unseen. These challenges prevent direct knowledge transfer across different units and over time, leading to the distribution gap between training and testing data and inducing performance degradation of those methods in real-world scenarios. To overcome this, our work introduces a novel approach for continuous test-time domain adaptation. This enables early-stage robust anomaly detection by addressing domain shifts and limited data representativeness issues. We propose a Test-time domain Adaptation Anomaly Detection (TAAD) framework that separates input variables into system parameters and measurements, employing two domain adaptation modules to independently adapt to each input category. This method allows for effective adaptation to evolving operating conditions and is particularly beneficial in systems with scarce data. Our approach, tested on a real-world pump monitoring dataset, shows significant improvements over existing domain adaptation methods in fault detection, demonstrating enhanced accuracy and reliability.

Employing language models to generate explanations for an incoming implicit hate post is an active area of research. The explanation is intended to make explicit the underlying stereotype and aid content moderators. The training often combines top-k relevant knowledge graph (KG) tuples to provide world knowledge and improve performance on standard metrics. Interestingly, our study presents conflicting evidence for the role of the quality of KG tuples in generating implicit explanations. Consequently, simpler models incorporating external toxicity signals outperform KG-infused models. Compared to the KG-based setup, we observe a comparable performance for SBIC (LatentHatred) datasets with a performance variation of +0.44 (+0.49), +1.83 (-1.56), and -4.59 (+0.77) in BLEU, ROUGE-L, and BERTScore. Further human evaluation and error analysis reveal that our proposed setup produces more precise explanations than zero-shot GPT-3.5, highlighting the intricate nature of the task.

Our goal is to $\textit{efficiently}$ discover a compact set of temporal logic rules to explain irregular events of interest. We introduce a neural-symbolic rule induction framework within the temporal point process model. The negative log-likelihood is the loss that guides the learning, where the explanatory logic rules and their weights are learned end-to-end in a $\textit{differentiable}$ way. Specifically, predicates and logic rules are represented as $\textit{vector embeddings}$, where the predicate embeddings are fixed and the rule embeddings are trained via gradient descent to obtain the most appropriate compositional representations of the predicate embeddings. To make the rule learning process more efficient and flexible, we adopt a $\textit{sequential covering algorithm}$, which progressively adds rules to the model and removes the event sequences that have been explained until all event sequences have been covered. All the found rules will be fed back to the models for a final rule embedding and weight refinement. Our approach showcases notable efficiency and accuracy across synthetic and real datasets, surpassing state-of-the-art baselines by a wide margin in terms of efficiency.

Large Language Models (LLMs) have shown human-like reasoning abilities but still face challenges in solving complex logical problems. Existing unidirectional chaining methods, such as forward chaining and backward chaining, suffer from issues like low prediction accuracy and efficiency. To address these, we propose a bidirectional chaining method, Bi-Chainer, which dynamically switches to depth-first reasoning in the opposite reasoning direction when it encounters multiple branching options within the current direction. Thus, the intermediate reasoning results can be utilized as guidance to facilitate the reasoning process. We show that Bi-Chainer achieves sizable accuracy boots over unidirectional chaining frameworks on four challenging logical reasoning datasets. Moreover, Bi-Chainer enhances the accuracy of intermediate proof steps and reduces the average number of inference calls, resulting in more efficient and accurate reasoning.