Nuclear power is a clean energy source that plays an important role in the development of the modern economy. The outdoor pipeline of nuclear power plant is an important part of connecting each plant and system, and if it leaks, it may lead to nuclear accidents and important losses of personnel and property. The polyethylene pipeline (PE pipeline) is a commonly used material for outdoor pipeline. To date, there is no automated process to detect impurities such as sand, hair, grass, and plastic in PE pipe joint welds. Impurities in joint welds can lead to serious long-term leakage. To improve the safety and reliability of the weld quality of outdoor PE pipelines in nuclear power plants, this study proposes a novel automated quality diagnosis method for outdoor PE pipelines in nuclear power plants based on a convolutional neural network (CNN) algorithm, which is used to improve the reliability and traceability of pipeline weld quality detection. This work used a large dataset of PE pipe joint samples created in harsh outdoor environments, in contrast to past studies that mainly used datasets created under controlled and laboratory conditions. The results show that the accuracy of the impurity recognition rate of this model is 86.29%, and that the automation process allows the identification of quality defects faster than the manual inspection method. The model will continue to collect samples and improve accuracy.

Although Artificial Intelligence (AI) models have demonstrated great success and efficiency in automatically countering online hate speech and misogyny, they suffer from a lack of explainability and transparency. Explainable Artificial Intelligence (XAI) is an apparent solution to make opaque black-box models more explainable, but remains under-explored for low resource and code-mixed languages. In this study, user-based evaluation of 3 explainability techniques for a misogyny classifier has been performed through simulatability, what-if explainability, and a qualitative feedback that covers several key dimensions. The participants were also required to mark the rationales to justify their predictions, and required to answer an open-ended question regarding their opinion on the explanations. Analysis of the results from 10 participants highlights that Integrated Gradients (IG) was more consistent and plausible than other techniques, and their preference for token attribution heatmaps and bar plots over Natural Language Explanations (NLE).

The rapid integration of generative artificial intelligence (AI) technologies into higher education raises urgent pedagogical and ethical questions, particularly around student engagement, academic integrity, and the future of learning. This study explores self-reported usage patterns, attitudes, and ethical perceptions related to generative AI among 87 undergraduate psychology students at a UK university. Using a cross-sectional online survey, we gathered data on students’ familiarity with generative AI, their attitudes towards it, their academic usage behaviours, and their justification (neutralisation) of its use in potentially inappropriate academic contexts. Over half the sample (54%) reported using generative AI to support their studies, primarily through ChatGPT. These users displayed significantly more positive attitudes towards AI and significantly higher neutralisation scores, indicating a stronger tendency to rationalise questionable or academically dishonest uses. Attitudes were a small but significant predictor of neutralisation. Most users reported using AI for concept explanation and idea generation, and the majority indicated they would use it again. Conversely, non-users more frequently endorsed restrictive views on AI use in assessments. The findings highlight a tension between the perceived pedagogical value of generative AI and the risks of academic dependency and erosion of critical thinking. Students recognise both the utility and the limitations of these tools, with many calling for clearer institutional guidance. We argue for a balanced approach that supports responsible AI integration into learning, paired with continued student-centred research to inform policy and practice.

Malware growth has accelerated due to the widespread use of Android applications. Android smartphone attacks have increased due to the widespread use of these devices. While deep learning models offer high efficiency and accuracy, training them on large and complex data sets is computationally expensive. Hence, a method that effectively detects new malware variants at a low computational cost is required. A transfer learning method to detect Android malware is proposed in this research. Because of transferring known features from a source model that has been trained to a target model, the transfer learning approach reduces the need for new training data and minimizes the need for huge amounts of computational power. We performed many experiments on 1.2 million Android application samples for performance evaluation. In addition, we evaluated how well our framework performed in comparison to traditional deep learning and standard machine learning models. In comparison to state-of-the-art Android malware detection methods, the proposed framework offers improved classification accuracy of 98.87%, a precision of 99.55%, recall of 97.30%, f1 measure of 99.42%, and a quicker detection rate of 5.14 ms by utilizing the transfer learning strategy.

In this work, a novel technique to train machine learning models is introduced, which is based on digital simulations of certain types of quantum systems. This represents a drastic departure from the standard approach which, to these days, is based on the use of actual physical quantum systems. Thus, to provide a clear context, a proper introduction to the field of quantum machine learning is first provided. Then, we proceed with a detailed description of our proposed method. To conclude, some preliminary, yet compelling, results are presented and discussed. Although at a seminal stage, the author firmly believes that this approach could represent a valid and robust alternative to the way machine learning models are trained today.

The field of Explainable AI (XAI) has focused primarily on algorithms that can help explain decisions and classification and help understand whether a particular action of an AI system is justified. These \emph{XAI algorithms} provide a variety of means for answering a number of questions human users might have about an AI. However, explanation is also supported by \emph{non-algorithms}: methods, tools, interfaces, and evaluations that might help develop or provide explanations for users, either on their own or in company with algorithmic explanations. In this article, we introduce and describe a small number of non-algorithms we have developed. These include several sets of guidelines for methodological guidance about evaluating systems, including both formative and summative evaluation (such as the self-explanation scorecard and stakeholder playbook) and several concepts for generating explanations that can augment or replace algorithmic XAI (such as the Discovery platform, Collaborative XAI, and the Cognitive Tutorial). We will introduce and review several of these example systems, and discuss how they might be useful in developing or improving algorithmic explanations, or even providing complete and useful non-algorithmic explanations of AI and ML systems.